nano materials

NANOMATERIALS

Edited by dr hab. inż. Lucyna Jaworska

Kraków 2010

Reviewers: Prof. dr hab. inż. Stanisław Wierzbiński dr hab. inż. Andrzej Dziadoń, prof PŚk Editorial board: dr hab. inż. Lucyna Jaworska Project of cover: dr inż. Paweł Kurtyka Text arrangement: dr inż. Paweł Figiel ISBN 978-83-912887-9-5 The monograph was supported by the Polish Ministry of Science and Higher Education as the Project DPN/N111/BIAŁORUŚ/2009 © Copyright by The Institute of Advanced Manufacturing Technology, Krakow 2010 Printed and cover from supplied materials by: Zakład Usług Poligraficznych, Krakow, J. Lea 114 Street Edition 100 copies

The main aim of this “Nanomaterials” monograph is to bring science and application together in nanoscale and nanostructured materials, with emphasis on modeling of new phenomena, synthesis, processing, characterization and application of materials containing nano-sized particles or nanostructures enabling novel properties and/or functions.

Nanotechnology may be thought of as an extension of traditional disciplines. In addition, traditional disciplines can be re-interpreted as specific applications of nanotechnology. A basic understanding of nanoscale phenomena and mechanisms in nanomaterials is still lacking. Nanomaterials are evolving and constantly appearing in new applications.

Each chapter of “Nanomaterials” presents the results of research conducted by institutions in Poland and abroad. Readers can become familiar with technologies and research methods which constitute original solutions. The chapters of this monograph emphasize original results relating to experimental, theoretical and computational applications of nanomaterials.

Editorial Board

dr hab inż Lucyna Jaworska

4

CONTENTS MATHEMATICAL MODEL OF ELECTROKINETIC PHENOMENA IN TWO–DIMENSIONAL CHANNELS 5 Vladimir V. Mityushev Natalia Rylko MECHANICAL PROPERTIES OF Si 3N4/SiC COMPOSITES WITH VARIOUS ADDITIONS 21 Piotr. Klimczyk, Lucyna Jaworska, Vladimir Urbanovich FRACTURE TOUGHNESS BEHAVIOUR OF TICNANO REINFORCED ALUMINA-ZIRCONIA TOOL COMPOSITES AT ELEVATED TEMPERATURES 45 Magdalena Szutkowska, Barbara Smuk, Marek Boniecki NANOCOMPOSITES IN THE ZrO 2-Al 2O3-WC SYSTEM 55 Zbigniew Pędzich, Wojciech Maziarz TEM INVESTIGATION OF MATRIX – SAFFIL TM FIBER INTERFACES IN ALUMINIUM ALLOYS BASED COMPOSITES 63 Jerzy Morgiel, Jacek Kaczmar, Małgorzata Pomorska, Krzysztof Naplocha TRIBOLOGICAL PROPERTIES OF SUPERSONIC SPRAYED NANOSTRUCTURED WC12CO COATINGS 74 Wojciech Żórawski AMORPHOUS Ti-Si-C THIN FILM DEPOSITED ON AISI 316L IN LOW TEMPERATURE 84 Agnieszka Twardowska IONIC METHODS OF COATINGS FORMATION FOR A SPECIAL APPLICATIONS 93 Bogusław Rajchel

Mathematical model of electrokinetic phenomena in two–dimensional channels

5

MATHEMATICAL MODEL OF ELECTROKINETIC PHENOMENA IN

TWO–DIMENSIONAL CHANNELS

VLADIMIR V. M ITYUSHEV1 NATALIA RYLKO

1

ABSTRACT Electroosmotic flows for a binary dilute electrolyte are theoretically studied in two–dimensional channels by expanding the solution into a series in terms of a dimensionless parameter related to the Debye–Huckel length. Analytical approximate formulae for the coupling coefficient are deduced.

1 Introduction

The study of electroosmotic flows through porous or dispersed media is important from a fundamental standpoint and for industrial applications. The external electric field E and the pressure gradient ∇p generate a macroscopic current density I and a Darcy seepage velocity u which for small E and ∇p are given by the linear formulae:

(1.1)

where ∇ is the gradient operator, K permeability, µ viscosity, σ is the electrical conductivity, α is the electroosmotic coupling coefficient. Theoretical and experimental study of the electrokinetic phenomena is given in [1]–[5]. In the present paper we pay attention to analytical formulas for the coupling coefficient α in the two-dimensional case. Consider a two-dimensional curvilinear channel bounded by the smooth dimension walls: 1 Department of Mathematics, Krakow Pedagogical University, ul. Podchorazych 2, 30-084 Krakow, Poland, E-mail: [email protected] d


6

(1.2)

The values of S±(x) are dimensionless and of order 1, i.e., S±(x)~1. For simplicity, sometimes a symmetric channel (S±(x)=±S(x)) is considered.

Introduce the dimension parameter and the dimensionless

parameter . Introduce two different dimensionless length scales: 1) Distorted coordinates

, (1.3)

Then (1.2) becomes

, - π < x < π (1.4)

where

2) Proportional coordinates

, (1.5)

Then (0.2) for symmetric channels becomes

, - π < x < π (1.6)

where S*(x*) = S(x). Further, the first dimensionless scale is used up to Sec. 4. The electrokinetic phenomena in the curvilinear 2D channel S-(x) < z < S+(x) is governed by the following PDE [1]

(1.7)

(1.8)

(1.9)

with the boundary conditions


7

(1.10)

(1.11)

(1.12)

(1.13)

Here, u = (u;w), is the outward normal derivative to the curves (1.2). Instead of the first equation (1.9) we consider equation

(1.14)

where . It is assumed that the functions S±(x) are twice continuously differentiable.

2 Dimensionless equations

Introduce the following dimensionless variables and parameters

(2.1)

where the most important in this paper the dimensionless parameter ν is introduced via the relation

(2.2)

Then

, (2.3)

Using (1.3)–(2.3) we rewrite the problem (1.9) in the dimensionless form. For shortness, primes are omitted

(2.4)


8

(2.5)

(2.6)

Hereafter, the subscripts x and z are used for the corresponding partial derivatives. The dimensionless parameter γi are introduced as

(i=1 ,2) (2.7)

Up to the end of this section dimension and dimensionless variables are again distinguished by primes and the boundary conditions (1.10)–(1.13) are rewritten in the dimensionless form. Consider equation of the walls in the dimensionless form (1.6). Then (1.10) becomes

(2.8)

u(x, S’±(x’)) = 0, w(x, S’±(x’)) = 0 (2.9)

g’(π, z) – g’ 9 (–π, z) 2πg0 (2.10)

i.e., the constant g0 has the form g0 = . Consider the boundary condition (0.13)

(i = 1, 2) (2.11)

where the normal vector n is determined by formula . Then (1.11) can be written in the form

on (i = 1, 2) (2.12)

Using (1.6) and (2.3) we have . Then (2.12) becomes


9

on z’ = S’±(x’) (i = 1, 2) (2.13)

3 Cascade on νννν2

We solve equations (2.4)–(2.6) with the boundary conditions (3.2)–(4.8) by expansion on ν2. Therefore, we actually assume that the

parameter is small, i.e., the ratio of the boundary layer κ−1 to the length of the periodicity cell L is small. This enables us to find the unknown functions in the form

(3.1)

Further in this paper, the zero order approximations from (3.1) are explicitly written. Omitting primes further we write the boundary condition (1.8) in the form

(3.2)

Substituting the first equation (3.1) into (2.4) and (3.2) and selecting the coefficients on the same powers of ν2 we obtain a cascade of the boundary value problem for .

The zeroth approximation of the problem for has the form

(3.3)

Solution to the problem (3.3) has the form

(3.4)

For symmetric channels when S+(x) = S(x) and S−(x) = − S(x), (3.4) takes the form

(3.5)

We have at the j–step , .

If is known, is easily obtained by the formula


10

(3.6)

where ψ(j)(x; z) is a partial solution to the ordinary differential

equation (3.6) on z. The functions are found from the boundary conditions (3.6). For instance, the first order approximation for symmetric channels has the form

(3.7)

It is possible to write an exact formula for for general channels, but it is too long.

4 Cascade for u

First, we introduce the auxiliary functions

(4.1)

Adding and subtracting equations (2.5) we obtain equations

(4.2)

(4.3)

(4.4)

where the dimensionless parameters δi are introduced via γi determined by (2.7)

(4.5)

Omitting primes further we write the boundary conditions (2.9), (2.10), (2.13) in the form

(4.6)

(4.7)

on z = S±(x) (i = 1, 2) (4.8)


11

The problem (4.2)–(4.4), (4.6)–(.8) in the zeroth approximation has the form

(4.9)

(4.10)

(4.11)

(4.12)

(4.13)

(4.14)

(4.15)

on z = S±(x) (i = 1, 2) (4.16)

Here, the function is known and has the form (3.4). One can consider (4.9)–(4.10) as a system of linear ordinary differential equations on the variable z of first order with respect to the functions

with zero boundary conditions (4.16). It is easily seen that this

system has only trivial solutions. This implies that the functions i depend only on x. Then it follows from (4.12) that , hence g(0) also depends only on x. Using (4.15) we obtain

(4.17)

Substitution of (4.17) into (4.11) yields

(4.18)

where is an undetermined function on x. Integrating (4.18) on z and using the relation (3.3) we obtain

(4.19)

where Cj(x) are undetermined functions (constant on z). Using the boundary conditions (4.14) for u(0) we obtain


12

(4.20)

Integrate equation (4.13) on z. Applying (4.19) we have

(4.21)

where C3(x) is undetermined function (constant on z),

Using the boundary conditions (4.14) for w(0) we have a formula for

C3(x) and a linear differential equation of first order for . Further constructive calculations are possible. However, the ultimate

explicit form of C3(x) and 1x is too huge. For symmetric channels when S+(x) = S(x) and S−(x) = −S(x), these formulas take a simpler form. The following formulas are valid for symmetric

channels, hence, C3(x) ≡ 0. The function satisfies equation

(4.22)

Its solution has the form

(4.23)

where C is an arbitrary constant. Hereafter, for shortness we write S

instead of S(x). The function has to be periodic with the period 2π. The latter condition yields

(4.24)

Therefore, the velocity in symmetric channels is properly defined in the zeroth approximation by formulas (4.19) and (4.21), where


13

(4.25)

has the form (3.23)–(3.24). Ultimately,

(3.26)

(3.27)

5 Verification of approximations

In the present section different dimensionless length scales noted by primes and stars are used. Consider the approximate solution (3.5) constructed by equation . This equation is considered as

an approximation of equation which is valid for sufficiently small ν. Therefore in order to check directly the

approximation (3.5), the value should be evaluated. However, it is not correct to estimate the ratio in the distorted variables (x’; z’). So, the value

(5.1)

is estimated in the coordinates (x*, z*) introduced by (1.5). We have

(5.2)

Then ∆ = ∆1 + ∆2, where


14

(5.3)

Therefore, (3.5) (or equivalently (5.2)) provides acceptable approximation if

∆1 < 1, ∆2 < 1 (5.4)

Inequalities (5.4) mean that the amplitude of the walls expressed by the first derivative and their curvature expressed by the second derivative are sufficiently small. Consider an example to clarify the conditions (5.4). Let the channel is restricted by the dimension walls

(5.5)

with small positive dimensionless parameter ε and given dimensionless coefficient λ which describes deviations of the walls. Equations (4.5) can be written in the dimensionless form

(5.6)

Using (5.3) calculate up to O(ε2)

(5.7)

Consider the cases: a) b’ = 1 (or equivalently bκ = 1). Then

(5.8)

b) b’ >> 1 (or equivalently bκ >> 1). Then

(5.9)

The relations (5.8)–(5.9) demonstrate geometric restrictions. For instance, if b ~ κ−1, (5.8) implies that has to be of order 1.


15

If b >> κ−1, (5.9) implies that despite b’ >> 1, it can be compensated by small ε or small λ. The proper expansion of by µ2 yields the proper order for u’, hence the proper order for the coupling coefficient discussed in the next section.

6 Coupling coefficient

The dimension coupling coefficient α is calculated by the integral [1]–[4]

(6.1)

where < f(x; z) > means the average of the function f(x; z) over the periodicity cell. Following [2]–[4] introduce the dimensionless

coefficient , where

(6.2)

The averaged of the dimensionless values can be calculated by formula

(6.3)

Here, is the dimensionless area of the periodicity cell

(6.4)

Applying the calculated zeroth approximations (3.4) and (4.19) for and u for symmetric channels we have in the zeroth

approximation


16

(6.5)

where C has the form (4.24). In the case of the straight channel S’(x’) = b’ formula (6.5) becomes the well–known formula [1]–[4]

(6.6)

Consider the permeability of the channel . Following [1]–[4] introduce the dimensionless permeability K = κ2K’ , which can be written in the form

(6.7)

Using (4.26) and (4.24) we calculate the zero approximation for the

symmetric channels< u’(0) >= . Dividing (6.2) by (6.7) we obtain

(6.8)

The following relation were discussed in [5]

(6.9)

where the dimension length scale Λ and the dimensionless length scale Λ’ = κ−1Λ are introduced by the following three ways:

(6.10)

Here, ∇ω is the local electric field in the channel filled by an electrolyte of constant conductivity and submitted to a macroscopic potential difference along the x–axis.


17

(6.11)

where F is the formation factor. F−1 is equal to the macroscopic conductivity of the channel filled by electrolyte of the normalized unit conductivity, K is the permeability.

(6.12)

where has the form (6.4), S’ is the length of the curve (1.6)

(6.13)

Consider the symmetric channel restricted by the walls

(6.14)

We are interested in the coupling coefficient and the permeability to O(ε2). It is convenient to find them in the form ,

, where and are the dimensionless coupling coefficient and the dimensionless permeability for the straight channel , the dimensionless constants α1 and k1 have to be found. First, we calculate and . For the straight channel, equations

(3.5) and (4.19) become , .

Then ; . Using (6.8) we get the relation

(6.15)

which coincides with (6.9) for Λ’ = b’. Recall that the normalized case ζ’ = 1, g0 = −1 is considered. The constant k1 were exactly calculated in [6] (formula (82) from [6] contains a slight error which is corrected below)


18

(6.16)

Equation (6.5) for the channel (6.14) becomes up to O(ε4)

(6.17)

Therefore,

(6.18)

Then (6.8) up to O(ε4) become

⋅

(6.19)

Figure 1: Results presented as functions on b for ε = 0:1 < b < 10. Data are for:

calculated by (6.8) dashed, with Λ1 calculated by (6.10) solid,

with Λ2 calculated by (6.11) dotted, with Λ3 calculated by (6.12) thin line coincided to (6.8).


19

Let λ denote the coefficient on ε2 of . It follows from (6.19) that

⋅

(6.20)

The corresponding coefficients on ε2 of the function

when Λ’ takes the values (5.10)–(5.12) have the form

(6.21)

(6.22)

(6.23)

The value λ from (5.20) does not coincide symbolically with the values given by (5.21)–(5.23). However, numerically one can see in the figure good agreement except the parameter Λ2 given by (5.11).

References [1] A. E. Malevich, V. V. Mityushev and P. M. Adler, Electrokinetic

phenomena in wavy channels. Journal of Colloid Interface Sci. 2010, 345(1), 72-87.

[2] D.Coelho, M.Shapiro, J.E.Thovert, P.M.Adler, Electroosmotic Phenomena in Porous Media, Journal of Colloid and Interface Science, 1996, 181, 169190.

[3] P.M.Adler, D.Coelho, J.-F.Thovert, M.Shapiro, Electrokinetic Phenomena in Porous Media and Around Aggregates, in Surfaces of nanoparticles and porous materials, ed. J.A. Schwarz, C.I. Contescu, Surfactant Science Series, (M. Dekker), 1999, 78, 211257

[4] P.M.Adler, Macroscopic Electroosmotic Coupling Coefficient in Random Porous Media Mathematical Geology, 2001, 33(1), 6393.


20

[5] A. Gupta , D. Coelho , P.M. Adler, Universal electro–osmotic formulae for porous media, Journalof Colloid Interface Sci. 2008 319(2), 549-554.

[6] A. E. Malevich, V. V. Mityushev and P. M. Adler, Stokes flow through a channel with wavy walls, Acta Mechanica, 2006, 182, 3-4, 151-182.

Mechanical properties of Si3N4/SiC composites with various additions

21

MECHANICAL PROPERTIES OF SI 3N4/SIC COMPOSITES WITH VARIOUS ADDITIONS

PIOTR KLIMCZYK 1, LUCYNA JAWORSKA 1,2,

VLADIMIR URBANOVICH 3

Abstract

Nano or submicro 30vol.%.Si3N4/70vol.%.SiC composites, with additions of Ti, TiB2 and cBN, were obtained by High Pressure - High Temperature (HPHT) sintering. Density, Young’s modulus, hardness, fracture toughness and coefficient of friction were measured. Microstructural (SEM) investigations and phase analysis (XRD) were also conducted for selected samples. Composites obtained from 30vol.% Si3N4/70vol.% SiC submicro powders with the addition of 30 vol.% cBN micro powder were characterized by the best mechanical properties.

1. Introduction

Silicon nitride has a favorable combination of properties that includes high strength over a broad temperature range, high hardness, moderate thermal conductivity, a low coefficient of thermal expansion, moderately high elastic modulus, and relatively high fracture toughness for a ceramic. Silicon nitride ceramics have reached large-scale production for cutting tools, bearings, turbocharger rotors and a variety of custom wear parts. Silicon carbide has many of the same applications as silicon nitride. Most silicon carbide materials have very high hardness (harder than alumina and silicon nitride) and thus have superior wear resistance [1, 2, 3, 4]. A major disadvantage of SiC ceramic materials is their low fracture toughness, which usually does not exceed about 3.5 MPa·m1/2 [5, 6]. 1 Mat. Eng. Dept., Inst. of Advanced Manuf. Techn., 37a Wroclawska St., 30-011 Krakow, Poland. 2 Dep. Technology, Pedagogical University, 2 Podchorazych St., 30-084 Krakow, Poland. 3 Scientific-Practical Materials Research Center of NAS of Belarus 19, P. Brovka St., Minsk 220072, Belarus.


22

In previous studies, attempts have been made to obtain an Si3N4 - SiC composite characterized by both high hardness and high fracture toughness [7, 8]. The purpose of the presented work is to study the effect of the addition of third-phase particles selected from metals (Ti) or ceramics (TiB2, cBN) to an Si3N4-70vol%SiC system on the mechanical properties of Si3N4-SiC composites obtained by the HPHT method. The main goal was to improve the fracture toughness of the investigated materials. When a particulate second phase is introduced into a brittle matrix, there are several toughening mechanisms that may be operative. E.g. metallic particles are capable of plastic deformation and thus absorption of energy and bridging of a growing crack, resulting in increased strengthening. (Fig. 1a) [9]. The addition of titanium particles into an Si3N4–SiC system may cause chemical reaction, and favorable, plastic, ceramic phase type titanium silicon nitride may appear. On the other hand, hard ceramic particles can introduce a favorable stress state, which can cause a toughening effect by crack deflection and crack bifurcation (Fig. 1b). Hard particles also improve the hardness and other mechanical properties of the composite as a whole.

a) b)

Fig. 1. Strengthening mechanisms in ceramic matrix composites with dispersed metallic (a) and ceramic (b) particles.

2. Experimental

2.1. Materials preparation

Powders used for the preparation of mixtures are listed in Table 1.


23

Table 1. Mixture composition.

Powder Granulation

[µm] Manufacturer Description

Si3N4(1) 0.6 H.C. Starck, Germany M11-grade (alpha>90%)

Si3N4(2) 1 – 5 AEE, US (alpha>85%)

Si3N4(08) 0.1 – 0.8 Goodfellow, UK (alpha)

SiC(2) 0.1 – 1 Goodfellow, UK (alpha)

SiC(3) - AGH, Poland specific surface 24 m2/g

TiH2(1) <44 Fluka, Switzerland

TiB2(4) 2.5 – 3.5 H.C. Starck, Germany F-grade

BN(M010) 0 – 0.01 Element6, South

Africa Micron+ABN (cubic)

BN(M36) 3 – 6 Element6, South

Africa Micron+ABN (cubic)

The following mixtures were prepared by mixing the appropriate powders in an isopropanol environment using a Fritsch Pulverisette 6 planetary mill.

A. 30Si3N4/70SiC composite:

A1) 30vol.% Si3N4(1) + 70 vol.% SiC(2) (submicro: Starck + Goodfellow)

A2) 30 vol.% Si3N4(08) + 70 vol.% SiC(2) (submicro: Goodfellow)

A3) 30 vol.% Si3N4(2) + 70 vol.% SiC(3) (micro: AEE + AGH)

A4) 30 vol.% Si3N4(2) + 70 vol.% SiC(2) (micro: AEE + submicro: Goodfellow)

B. 30 Si3N4/70SiC composite + Ti:


24

B1) 28 vol.% Si3N4(1) + 64 vol.% SiC(2) + 8 vol.% Ti - from TiH2(1) (submicro: Starck + Goodfellow + micro: Fluka)

C. 30 Si3N4/70SiC composite + TiB2:

C1) 28 vol.% Si3N4(1) + 64 vol.% SiC(2) + 8 vol.% TiB2(4) (submicro: Starck + Goodfellow + micro: Starck)

C2) 21 vol.% Si3N4 (1)+49 vol.% SiC(2)+30 vol.% TiB2(4) (submicro: Starck + Goodfellow + micro: Starck)

D. 30 Si3N4/70SiC composite + cBN:

D1) 28 vol.% Si3N4 (1) + 64 vol.% SiC(2) + 8 vol.% cBN(M36) (submicro: Starck + Goodfellow + micro: E6)

D2) 28 vol.% Si3N4(08) + 64 vol.% SiC(2) + 8 vol.% cBN(M010) (submicro: Goodfellow + nano: E6)

D3) 21 vol.% Si3N4(1) + 49 vol.% SiC(2) + 30 vol.% cBN(M36) (submicro: Starck + Goodfellow + micro: E6)

After drying, the mixtures were preliminarily compressed into pellets of diameter 15 mm and height 5 mm under pressure of ~200 MPa. The green bodies with the addition of TiH2 were additionally annealed in a vacuum at a temperature of 800oC for 1h in order to remove the hydrogen and obtain pure metallic titanium. The composites were obtained at high pressure (6 GPa) in the temperature range of 790 – 2030 °C using a Bridgman-type toroidal apparatus (Fig. 2). The sintering temperatures were established experimentally for each composite to obtain crack-free samples with the highest values of density and mechanical properties. Duration of the sintering process was 60 s.


25

Fig. 2. Sintering process in a Bridgman-type HPHT system. Quasi-hydrostatic compression of the preliminary consolidated powders (sample - 1) is achieved as a result of plastic deformation of the gasket material (2) between anvils (3); electrical heating is provided by a high-power transformer (4) and graphite resistive heater (5).

The sintered compacts were subsequently ground to remove remains of graphite after the technological process of sintering and to obtain the required quality and surface parallelism for physical and mechanical studies.

2.2. Methods of investigation

Densities of the sintered samples were measured by the hydrostatic method. The uncertainty of the measurements was below 0.02 g/cm3, which gave a relative error value of below 0.5 % (excluding measurements of small pieces of broken samples, where error was up to 0.1 g/cm3, due to their insufficient volume and mass). Young’s modulus of the samples obtained by HPHT sintering was measured based on the transmission velocity of ultrasonic waves through the sample, using a Panametrics Epoch III ultrasonic flaw detector. Calculations were carried out according to the following formula:


26

22

222 43

=TL

TLT CC

CCCρE (1)

where: E - Young’s modulus, CL - velocity of the longitudinal wave, CT - velocity of the transversal wave, ρ - density of the material.

The velocities of transverse and longitudinal waves were determined as a ratio of sample thickness and relevant transition time. The accuracy of calculated Young’s modulus (Eq. 1) was estimated to be below 2 %. Hardness of selected samples was determined by the Vickers method using a digital Vickers Hardness Tester (FUTURE-TECH FV-700). Five hardness measurements, with indentation loads of 2.94, 9.81 and 98.1 N, were carried out for each sample. Standard deviations of HV values were relatively high but usually no more than 5 % of the average values. Indentation fracture toughness was calculated from the length of cracks which developed in a Vickers indentation test (with indentation load - 98.1 N) using Niihara’s equation (Eq. 2):

2

35

2

1290−

a

c.=

E

H

aH

K IC

ϕϕ

(2)

where: KIC - critical stress intensity factor, ϕ - constrain factor, H - Vickers hardness, E - Young’s modulus, a - half of indent diagonal, c - length of crack.

Microstructural observations were carried out on the densified materials using a JEOL JXA-50A Scanning electron Microscope equipped with back scattering electron (BSE) imaging. In ball-on-disc tests, the coefficients of friction for contact with an Si3N4 ball were determined using a CETR UMT-2MT (USA) universal mechanical tester. In the ball-on-disc method, sliding contact is produced by pushing a ball specimen onto a rotating disc specimen under a constant load. Tests were carried out without lubricant. The loading mechanism applied a controlled load Fn to the ball holder


27

(Fig. 3) and the friction force was measured continuously during the test using an extensometer. For each test, a new ball was used or the ball was rotated such that a new surface was in contact with the disc. After mounting of the ball and sample, materials were washed in ethyl alcohol and dried.

Fig. 3. Material pair for the ball-on-disc method: 1 – Si3N4 ball; 2 – sample (disc)

The size of the disc-shaped samples was ~13.5 x 3.8 mm; the surface of the discs flat and parallel to within 0.02 mm; and the roughness of the test surface not more than 0.1 µm Ra. The test samples were ground using diamond wheels and polished using diamond slurries. The following test conditions were established:

• ball diameter: 2 mm,

• applied load: 4 N,

• sliding speed: 0.1 m/s,

• diameter of the sliding circle: 2 - 5 mm,

• sliding distance: 100 m,

• calculated duration of the test: 1000 s.

Friction coefficient was calculated from the following equation:

Fn

F f=µ (3)

where: Ff is the measured friction force, and Fn is the applied normal force.


28

4. Results and discussion

Density and Young’s modulus of various 30 vol.% Si3N4/70 vol.% SiC composites sintered at different temperatures with and without additions of Ti, TiB2 and cBN phases are presented in Figs. 4-15 and in Table 2.

2,902,922,942,962,983,003,023,043,063,083,103,123,143,163,183,203,22

14

00

15

00

16

00

17

00

18

00

19

00

20

00

21

00

Sintering temperature, oC

De

nsity

, g/c

cm

Theoretical 30Si3N4/70SiC

A1) 30Si3N4(1)+70SiC(2)(submicro:Starck+Goodfellow)

A2) 30Si3N4(08)+70SiC(2) (submicro:Goodfellow)

A3) 30Si3N4(2)+70SiC(3) (micro:AEE+AGH)

A4) 30Si3N4(2)+70SiC(2)(micro:AEE+submicro:Goodfellow)

Fig. 4. Density of various kinds of 30 Si3N4/70 SiC composites sintered at different temperatures. Dark symbols – samples without cracks; white symbols – samples with cracks; white symbols lying on temperature axis – broken samples.


29

200

220

240

260

280

300

320

340

360

380

400

420

140

0

150

0

160

0

170

0

180

0

190

0

200

0

210

0


Yo

un

g m

odu

lus,

GP

a

Theoretical 30Si3N4/70SiC

A1) 30Si3N4(1)+70SiC(2)(submicro:Starck+Goodfellow)

A2) 30Si3N4(08)+70SiC(2) (submicro:Goodfellow)

A3) 30Si3N4(2)+70SiC(3) (micro:AEE+AGH)

A4) 30Si3N4(2)+70SiC(2)(micro:AEE+submicro:Goodfellow)

Fig. 5. Young’s modulus of various kinds of 30 Si3N4/70 SiC composites sintered at different temperatures. Dark symbols – samples without cracks; white symbols – samples with cracks; white symbols lying on temperature axis – broken samples.

3,10

3,12

3,14

3,16

3,18

3,20

3,22

3,24

3,26

3,28

3,30

3,32

3,34

70

0

80

0

90

0

10

00

11

00

12

00

13

00

14

00

15

00

16

00

17

00

18

00

19

00

20

00


De

nsi

ty, g

/ccm

Theoretical 30Si3N4/70SiC +8Ti

B1)28Si3N4(1)+64SiC(2)+8Ti(1)(submicro:Starck+Goodfellow+micro:Fluka)

Fig. 6. Density of Si3N4/70 SiC + 8 vol.% Ti composites sintered at different temperatures. Dark symbols – samples without cracks; white symbols – samples with cracks; white symbols lying on temperature axis – broken samples.


30

100120140160180200220240260280300320340360380400

70

0

80

0

90

0

100

0

110

0

120

0

130

0

140

0

150

0

160

0

170

0

180

0

190

0

200

0


Yo

un

g m

odu

lus,

GP

a

Theoretical 30Si3N4/70SiC +8Ti

B1)28Si3N4(1)+64SiC(2)+8Ti(1)(submicro:Starck+Goodfellow+micro:Fluka)

Fig. 7. Young’s modulus of Si3N4/70SiC + 8 vol.% Ti composites sintered at different temperatures. Dark symbols – samples without cracks; white symbols – samples with cracks; white symbols lying on temperature axis – broken samples.

3,16

3,18

3,20

3,22

3,24

3,26

3,28

3,30

3,32

16

00

17

00

18

00

19

00


De

nsi

ty, g

/ccm

Theoretical 30Si3N4/70SiC +8TiB2

C1)28Si3N4(1)+64SiC(2)+8TiB2(4)(submicro:Starck+Goodfellow+micro:Starck)

Fig. 8. Density of Si3N4/70SiC + 8 vol.% TiB2 composites sintered at different temperatures. Dark symbols – samples without cracks; white symbols – samples with cracks; white symbols lying on temperature axis – broken samples.


31

340

350

360

370

380

390

400

410

420

430

16

00

17

00

18

00

19

00


Yo

un

g m

od

ulu

s, G

Pa



Fig. 9. Young’s modulus of Si3N4/70SiC + 8 vol.% TiB2 composites sintered at different temperatures. Dark symbols – samples without cracks; white symbols – samples with cracks; white symbols lying on temperature axis – broken samples.

3,40

3,42

3,44

3,46

3,48

3,50

3,52

3,54

3,56

3,58

3,60

3,62

16

00

17

00

18

00

19

00

20

00

21

00

22

00


De

nsi

ty, g

/ccm



Fig. 10. Density of Si3N4/70SiC + 30 vol.% TiB2 composites sintered at different temperatures. Dark symbols – samples without cracks; white symbols – samples with cracks; white symbols lying on temperature axis – broken samples.


32

320

340

360

380

400

420

440

460

1600

1700

1800

1900

2000

2100

2200


You

ng m

odul

us, G

Pa

Theoretical 30Si3N4/70SiC+ 30TiB2


Fig. 11. Young’s modulus of Si3N4/SiC + 30 vol.% TiB2 composites sintered at different temperatures. Dark symbols – samples without cracks; white symbols – samples with cracks; white symbols lying on temperature axis – broken samples.

2,90

2,95

3,00

3,05

3,10

3,15

3,20

3,25

160

0

170

0

180

0

190

0

200

0


De

nsi

ty, g

/ccm

Theoretical 30Si3N4/70SiC +8BN

D1)28Si3N4(1)+64SiC(2)+8BN(M36)(submicro:Starck+Goodfellow+micro:E6)

D2)28Si3N4(08)+64SiC(2)+8BN(M010)(submicro:Goodfellow+nano:E6)

Fig. 12. Density of Si3N4/70 SiC + 8 vol.% cBN composites sintered at different temperaturesDark symbols – samples without cracks; white symbols – samples with cracks; white symbols lying on temperature axis – broken samples.


33

320

340

360

380

400

420

440

460

1600

1700

1800

1900

2000


You

ng m

odul

us, G

Pa

Theoretical 30Si3N4/70SiC+ 8BN


D2)28Si3N4(08)+64SiC(2)+8BN(M010)(submicro:Goodfellow+nano:E6)

Fig. 13. Young’s modulus of Si3N4/70SiC + 8 vol.% cBN composites sintered at different temperatures. Dark symbols – samples without cracks; white symbols – samples with cracks; white symbols lying on temperature axis – broken sample.

3,25

3,30

3,35

3,40

3,45

3,50

3,55

3,60

3,65

3,70

3,75

3,80

3,85

3,90

3,95

16

00

17

00

18

00

19

00

20

00


De

nsi

ty, g

/ccm

Theoretical 30Si3N4/70SiC +30BN


Fig. 14. Density of Si3N4/70SiC + 30 vol.% cBN composites sintered at different temperatures. Dark symbols – samples without cracks; white symbols – samples with cracks; white symbols lying on temperature axis – broken sample.


34

420

440

460

480

500

520

540

560

580

1600

1700

1800

1900

2000


You

ng m

odul

us, G

Pa

Theoretical 30Si3N4/70SiC+ 30BN


Fig. 15. Young’s modulus of Si3N4 /70SiC + 30 vol.% cBN composites sintered at different temperatures. Dark symbols – samples without cracks; white symbols – samples with cracks; white symbols lying on temperature axis – broken sample.


35

Table 2 /part 1. Physical-mechanical properties of the best samples selected from different modifications of Si3N4/SiC composite.


36

Table 2 /part 2. Physical-mechanical properties of the best samples selected from different modifications of Si3N4/SiC composites.


37

Among the composites sintered without additional phases, the highest degree of densification and best mechanical properties were demonstrated by composite A1, obtained from submicron powders 30 vol.% Si3N4(1) + 70 vol.% SiC(2) (submicro: Starck + Goodfellow) (Figs. 4, 5 and Table 2). This composite was selected for modification by the addition of third phase particles. Modification of the 30 vol.% Si3N4/70 vol.% SiC composite by the addition of Ti was not successful. Samples from series B1, with the addition of 8% Ti introduced in the form of TiH2, sintered at low temperatures, were characterized by a very low Young's modulus, whilst all the samples sintered at temperatures above ~1200 °C were cracked. A decrease in density was observed with increasing sintering temperature, whilst Young's modulus showed an upward trend (Figs. 6 and 7). Composites with the addition of TiB2 were characterized by a high degree of densification, a high Young’s modulus and improved KIC as compared to the unmodified composite. No improvement in hardness was observed (Table 2). In the case of material C1, with the addition of 8% TiB2, there appears to be some increase in density and Young’s modulus with increasing temperature, (Figs. 8-9). Composite C2, with the addition of 30 vol.% TiB2, shows an increase in density with sintering temperature up to a maximum value, and then its stabilization. Further increase of the sintering temperature results in cracking of the samples. This is typical behavior in the sintering of advanced ceramics by the HPHT method (Fig. 10). The Young's modulus of this material decreases with increasing sintering temperature, which is not typical (Fig. 11). Composites from series D3, modified by the addition of 30 vol.% cBN micropowder, showed the highest degree of densification, equal to 118% of theoretical density - EDS analysis showed a high content of tungsten carbide and zirconium dioxide from the vessel and grinding media used to prepare the mixtures - the best mechanical properties (Table 2) and the lowest percentage of cracked samples. (Figs. 14 and 15).


38

SEM microstructures of 30Si3N4/70SiC with and without the addition of TiB2 and cBN are presented in Fig. 16.

A1) 30 vol.% Si3N4(1) + 70SiC(2) (submicro: Starck + Goodfellow).

A3) 30 vol.% Si3N4(2) + 70 vol.% SiC(3) (micro: AEE + AGH).


39

C1) 28 vol.% Si3N4(1) + 64 vol.% SiC(2) + 8 vol.% TiB2(4)

(submicro: Starck+Goodfellow + micro: Starck).

C2) 21 vol.% Si3N4 (1)+49 vol.% SiC(2)+30 vol.% TiB2(4)

(submicro: Starck + Goodfellow + micro: Starck.


40

D1) 28 vol.% Si3N4 (1) + 64 vol.% SiC(2) + 8 vol.% cBN(M36)

(submicro: Starck + Goodfellow + micro: E6).

D3) 21 vol.% Si3N4(1) + 49 vol.% SiC(2) + 30 vol.% cBN(M36)

(submicro: Starck + Goodfellow + micro: E6).

Fig. 16. SEM microstructures of selected 30Si3N4/70SiC composites with and without the addition of TiB2 and cBN.

The microstructures of investigated samples are compact and dense, with the ingredients uniformly distributed in the volume of composite. This demonstrates successful blending, using a planetary mill; EDS analysis, however, showed a high content of tungsten carbide and zirconium dioxide from the vessel and grinding media


41

used to prepare the mixtures (white areas visible in the microstructures (Fig. 16). Three exemplary curves representing measurements of coefficient of friction for sample C1: 30 vol.% Si3N4/70 vol.% SiC + 8 vol.% TiB2, and the mean curve, are presented in Fig. 17. A comparison of mean curves of friction coefficient for various kinds of 30 vol.% Si3N4/70 vol.% SiC composites, with and without the addition of cBN and TiB2, is presented in Fig. 18.

0 100 200 300 400 500 600 700 800 900 10000,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1,0 C1 30Si3N4/70SiC (submicro) + 8TiB2_Test1 C1 30Si3N4/70SiC (submicro) + 8TiB2_Test2 C1 30Si3N4/70SiC (submicro) + 8TiB2_Test3 C1 30Si3N4/70SiC (submicro) + 8TiB2_Mean

µ [-

]

Time [s]

Fig. 17. Coefficient of friction of 30 vol.% Si3N4/70 vol.% SiC + vol.% 8TiB2 composite.


42

0 100 200 300 400 500 600 700 800 900 10000,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1,0 A1 30Si3N4/70SiC (submicro) A4 30Si3N4/70SiC (micro/submicro) C1 30Si3N4/70SiC (submicro) + 8TiB2 C2 30Si3N4/70SiC (submicro) + 30TiB2 D3 30Si3N4/70SiC (submicro) + 30cBN

µ [-

]

Time [s]

Fig. 18. Coefficient of friction of selected 30 vol.% Si3N4/70 vol.% SiC composites with and without the addition of TiB2 and cBN (mean curves).

High coefficients of friction generate thermal stress, which is detrimental to the wear behavior of materials. Hard ceramic bodies – possessing high fracture toughness and low coefficients of friction – used in mechanical systems that involve high loads, velocities and temperatures, will reduce costs and be less harmful to the environment. Obtaining coefficients of friction below 0.2 is unlikely since, under non-lubricated conditions, current dual-phase ceramics typically have coefficients of friction of 0.5–0.8 [10]. Unmodified composites had the highest coefficients of friction. Average values of friction coefficient for materials A1 and A4 were 0.60 and 0.56 respectively. Composites modified by the addition of a TiB2 phase were characterized by intermediate values of friction coefficient. Average values of friction coefficient for materials C1 and C2 were 0.48 and 0.46 respectively. The composite with the addition of 30% cBN was characterized by the lowest average coefficient of friction, at only 0.36.


43

Summary

Various modifications of 30Si3N4/70SiC (volume ratio) composites, with additions of Ti, TiB2 and cBN, obtained by HPHT sintering, were investigated. Composites modified by the addition of Ti show deterioration of properties in comparison with the unmodified composites. Composites modified by the addition of TiB2 show only a partial improvement of properties in comparison with the unmodified composites. Composites modified by the addition of 30% cBN micropowder are characterized by the best combination of Young’s modulus, hardness, fracture toughness and coefficient of friction. Such properties predispose 30Si3N4/70SiC + 30% cBN composites to various advanced engineering applications. Wear tests and cutting tests (intended in the future) will show the range of applications of this material in machining.

Acknowledgements

This study was carried out within the framework of the project funded by the Polish Ministry of Science and Higher Education (Project number: DPN/N111/BIALORUS/2009).

REFERENCES

[1] Peng H., Spark Plasma Sintering of Si3N4-Based Ceramics, Doctoral Dissertation, Department of Inorganic Chemistry, Stockholm University 2004.

[2] Pierson H.O., Handbook of refractory carbides and nitrides, Noyes Publications, Westwood, New Jersey 1996.

[3] Richerson D.W., Advanced ceramic materials, in J. K. Wessel (ed.), Handbook of advanced materials, John Wiley & Sons, Inc., Hoboken, New Jersey 2004.

[4] Eblagon F., Ehrle B., Graule T., Kuebler J., Development of silicon nitride/silicon carbide composites for wood-cutting tools, Journal of the European Ceramic Society, 2007, 27, s.419–428.

[5] Lee S.M., Kim T.W., Lim H.J., Kim C., Kim Y.W., Lee K.S., Mechanical Properties and Contact Damages of Nanostructured Silicon Carbide


44

Ceramics, Journal of the Ceramic Society of Japan, 2007, 15 (5), s. 304-309.

[6] Suyama S., Kameda T., Itoh Y., Development of high-strength reaction-sintered silicon carbide, Diamond and Related Materials, 2003, 12, s. 1201-1204.

[7] Klimczyk P., Urbanovich V.S., Micro-, submicro- and nano-Si3N4 – SiC composites sintered by the HPHT method, Archives of Materials Science and Engineering, 2009, 39/2, s. 92-96.

[8] Klimczyk P., Mechanical properties of Si3N4-SiC composites sintered by the HPHT method, Proceedings of CIMTEC 2010 - 12th International Ceramics Congress – in press.

[9] Yeomans J.A., Ductile particle ceramic matrix composites - Scientific curiosities or engineering materials?, Journal of the European Ceramic Society, 2008, 28, s.1543–1550.

[10] Kerkwijk B., Garcia M., Zyl W.E., Winnubst L., Mulder E.J., Schipper D. J., Verweij H., Friction behaviour of solid oxide lubricants as second phase in α-Al 2O3 and stabilised ZrO2 composites, Wear 2004, 256, s.182–189.

Fracture toughness behaviour of TiCnano reinforced alumina-zirconia tool composites at elevated temperatures

45

FRACTURE TOUGHNESS BEHAVIOUR OF TIC NANO REINFORCED ALUMINA-ZIRCONIA

TOOL COMPOSITES AT ELEVATED TEMPERATURES

MAGDALENA SZUTKOWSKA

1, BARBARA SMUK1,

MAREK BONIECKI2

Abstract

Recent advances in high-speed cutting materials have focused on reinforcing alumina with different carbides, oxides and nitrides in order to improve hardness, fracture toughness and wear resistance. The potential use of this type of material for cutting tool applications has not yet been determined. The present study reports selected properties obtained by reinforcing Al2O3 – 10 mass% ZrO2 (partially stabilized with Y2O3 -Y5) composite with TiC in amount of 5 mass%. Specimens were prepared based on submicro- and nano-scale trade powders. Vicker’s hardness (HV) and fracture toughness (KIC) at room and elevated temperatures characteristic for tool work were evaluated. Wear resistance (Vn), a very important property for tool ceramics, was determined by speed of mass loss. Microstructural observations of the specimens were carried out using a scanning electron microscopy (SEM). The addition of TiC nanopowders does not result in a significant improvement in fracture toughness. Preliminary industrial tests confirm the higher cutting performance of TiCsubmicro reinforced alumina-zirconia tool composites in comparison to the same composites reinforced with TiCnano.

1. Introduction

Engineered ceramics are used in thermal and structural applications requiring high temperature resistance, high hardness and chemical inertness. Applications that exploit the thermal structural properties of ceramics commonly include cutting tool inserts, wear resistant components, ballistic armour, heat exchangers, burner tubes, 1 Mat. Eng. Dept., Inst. of Advanced Manuf. Techn., 37a Wroclawska St., 30-011 Krakow, Poland. 2 Dep. of Cer. and Joints., Inst. of Elec. Mat. Techn., 133 Wolczynska St., 01-919 Warsaw, Poland


46

prosthetics, dental implants, heat engine components and thermal barrier coatings. The main characteristics limiting these ceramics in the above and other engineering applications are their lack of plastic behaviour at room temperature and their low tolerance to flaws, i.e. low fracture toughness, which lead to catastrophic failure. Reinforcing ceramics with particles, whiskers, platelets, continuous fibres and discontinuous fibres significantly improves their strength, toughness and apparent ductility [1]. In ceramic composites, reinforcement usually increases strength indirectly by increasing the toughness of the matrix. Composite design can also be used to tailor other important properties such as high-temperature strength and thermal shock resistance, wear resistance, friction, thermal and electrical conductivity and thermoelastic properties. The need for high-temperature, oxidation-resistant materials has driven research into oxide-oxide composites. By an appropriate adjustment of the concentration of its components, alumina matrix ceramic composites (AMCC) can offer a high level of strength properties, toughness and hardness [2]. Previous studies have indicated that the toughening improvement by the addition of a dispersed second phase is based on several mechanisms, such as transformation toughening, crack deflection, thermal residual stresses, microcrack formation, debonding, pull-out, and crack bridging [3]. In an Al 2O3 matrix, t-ZrO2 grains undergo the t→m transformation (stress induced phase transformation) and microcracks form around pretransformed m-ZrO2 grains (microcrack formation), these are the major toughening mechanisms in Al2O3-ZrO2 composites. The mechanism of this process based on the polymorphic martensitic transformation of the t-ZrO2

tetragonal phase into an m-ZrO2 monoclinic phase during cooling from sintering temperature to room temperature, - allows an increase in the strength and/or fracture toughness of Al2O3-ZrO2

composites [4]. Optimization of the stability of tetragonal zirconia can lead to improved mechanical properties in zirconia-toughened alumina composites (ZTA). It is suggested that the highest toughness can be obtained without a reduction in hardness if the additives are adjusted to optimize the mixture of zirconia phases of different sizes (nano and submicro scales). On the other hand, pure titanium carbide (TiC) has many attractive properties, such as high


47

hardness, low density, and relatively high thermal and electrical conductivity [5]. These ceramic composites, to their great thermodynamic stability, high hardness, and compression strength of the TiC additive, have a better cutting properties in comparison to oxide ceramics, and can be used for precision machining of hard work-pieces. The advantages of such tools are the possibility of high duty working, obtaining of surfaces with very low roughness (in many cases grinding is eliminated) and the possibility of ecological “dry cutting” without the use of cutting fluids. A comparison of cutting properties for different tool ceramics is presented in Fig.1 [6].

Fig.1. Comparison of cutting properties for different tool ceramics [6]

Thus the main aim of this study is to obtain new titanium carbide-reinforced alumina-based composite tool ceramics that combine excellent toughness with increased hardness. These properties result from the addition to zirconia-toughened alumina ceramics of TiC particles in submicro and nano scale.

2. Experimental Procedure

Alumina, zirconia partially stabilized with yttria, non-stabilized zirconia and titanium carbides were used as raw materials to manufacture the titanium carbide-reinforced alumina-based composites. A commercial Alcoa alumina powder (containing 85.0% α-phase, of 99.8% purity) type A16SG with a submicron


48

particle size of below 0.5 µm and a nano particle size of 40 nm, with average size of agglomerates of 150 nm produced by Inframat Advanced Materials, USA were used to prepare the tested composites. Experiments were carried out on alumina-based composites with the addition of: 10 mass% ZrO2 with modification of the zirconia phase (partially stabilized with 5% mass Y2O3 –PYT05.0 (ZY5) produced by Unitec Ceramics, England, with an average particle size of 0.9-1.1 µm; ZrO2 stabilized with 3% mol. Y2O3 produced by Inframat Advanced Materials, USA (marked as YSZnano) and a monoclinic phase of zirconia in submicron (m-ZrO2) and nano scale (m-ZrO2nano) produced by Fluka, Germany. A commercial TiC powder with different particle sizes as a standard TiC (grade ST120 produced by H.C. Starck, Germany) with an average particle size of 1.0- 4.0 µm and TiCnano with agglomerates of : ≤130 nm and ≤200 nm produced by Aldrich Sigma, Germany. Sintering additives, such as MgOnano (0.3 mass%), were introduced to inhibit grain growth. The initial compositions of compounds selected for testing are presented in Table 1. Table 1. Composition of selected compounds

Compound composition [mass%] Al 2O3 ZY5 ZrO2

(m) TiC Compound

[µm] [nm] [µm] [nm] [µm] [nm] [µm] <130 nm

<200 nm

B 85.0 5.0 5.0 5.0 B1 42.5 42.5 5.0 5.0 5.0 B2 42.5 42.5 5.0 5.0 5.0

Components were mixed for approximately thirty hours in alumina mills with zirconia balls, with the addition of a plasticizer. Materials, uniformly set, after plasticizing and drying, were granulated. Green compacts with dimensions of 5.5 ×3.0 ×35 mm and 16.5x16.5x10 mm were uniaxially pressed at 100 MPa and then cold isostatically pressed at 200 MPa. Ceramic composites were sintered at a maximum temperature of 1973 K at constant heating and cooling rates of the furnace. Sintering of the specimens was carried out in a high-temperature following measurements were


49

performed on the tested specimensVickers hardness HV1, fracture toughness at room temperature KIC (based on 3PB), fracture toughness at an elevated temperature of 873K KIC(ET), wear resistance Vn (determined by speed of mass loss), apparent density ρp, and porosity Pc (determined by the hydrostatic method). Single Edge Notched Beam (SENB) specimens (mechanically notched) with dimensions 1.5×4.0×35.0±0.1mm were used to determine fracture toughness by means of a conventional method based on three-point bending of specimens (3PB). An initial 0.9 mm deep notch was produced a diamond saw (thickness 0.20 mm) and then the notch tip was pre-cracked with a thin diamond saw (thickness 0.025 mm). The total initial notch length was approximately 1.1 mm. The relationship KIC = f (c) is given by the equations (1, 2) [7]:

2/12

5.1 YcBW

SPK c

IC = (1)

( )

−+−

Π= ∑=

4

0,

)1(3738.01

32

ji

ji

ij S

WAY βββ

β (2)

where: Pc = critical load, S = support span, W= width, B = specimen thickness, Y = a geometric function, c = crack length, β = c/W and Aij is the coefficients given by Fett [8]. Measurements of fracture toughness at an elevated temperature of 873K were carried out on a ZWICK 1446 instrument with mounted electrical furnace (Fig. 2). Modulus of elasticity (Young’s modulus) of the tested composites was determined by a measuring the transmission velocity of longitudinal and transversal ultrasonic waves through the sample.


50

Fig.2. Measuring position for determination of fracture toughness at elevated

temperatures. Probe sets work together with a Panametrics Epoch III ultrasonic flaw detector connected to a controlling computer. Calculations were made using the following formula:

22

222 43

TL

TLT

CC

CCCE

−−

= ρ (4)

where: E-Young’s modulus, CL -velocity of the longitudinal wave, CT -velocity of the transversal wave, ρ- density of the material. The velocities of transversal and longitudinal waves were determined as a ratio of sample thickness and relevant transition time. The accuracy of calculated Young’s modulus from equation (4) was estimated to be below 2 %. Wear resistance was determined by a method based on the measurement of mass decrement rate during wear of the specimen against an SiC80 abrasive cloth. The following formula was used to calculate mass decrement rate (Vn):

Vn=TF

m

P ⋅⋅∆⋅

ρ1000

[µm/h] (5)


51

where: ∆m-absolute mass wear, ρp-apparent density, F-specimen contact surface, T-time. Microstructural observations of the specimen were carried out using a JEOL JSM-6460LV scanning electron microscope. X-ray diffraction was used both to identify phases and to assess the ratio of tetragonal zirconia phase (t) to monoclinic zirconia phase (m). Preliminary cutting tests in the turning of constructional alloy steel (40H) of hardness 252 HB were carried out for SNGN 12 04 08 type inserts. Tool life (Tmean) of the insert cutting edge was measured for tested composite inserts at the following cutting parameters: speed (vc) = 150 m/min, feed (f) = 0.15 mm/rev., depth (ap) = 0.5 mm, wear criterion (width of flank wear land) VBB = 0.30 mm. Cutting tests were performed without cooling lubricant.


The results obtained from the tests concerning: mechanical properties, apparent density ρp, and porosity Pc of tested ceramic composites of various compositions are presented in Table 2. Table 2. Physical and mechanical properties of tested ceramic composites: apparent density ρp, porosity Pc, Young’s modulus E, Vickers hardness HV1, fracture toughness at room temperature KIC and elevated temperature (873K) KIC(ET), wear resistance Vn.

Materials

Appa-rent

density ρp

[g/cm3]

Poro-sity Pc

[%]

Vickers hardness

HV1 [GPa]

Young’s modulus

E [GPa]

Critical stress

intensity factor KIC

[MPa m1/2]

Critical stress

intensity factor KIC(ET)

[MPa m1/2]

Wear resistance

Vn [µm / h]

B 4.13 2.3 18.8 378 4.6 (0.1) 4.3 (0.04) 5.7 B1 4.11 2.7 18.4 376 4.5 (0.2) 4.4 (0.4) 6.8 B2 4.09 3.5 18.3 358 4.2 (0.2) 4.0 (0.3) 7.8 Tested titanium carbide-reinforced alumina-based composite tool ceramics exhibited: high Vickers hardness (in the range 18.3-18.8 GPa), similar values of critical stress intensity factor KIC at room temperature (in the range 4.2-4.6 MPa m1/2), KIC(ET) at an elevated temperatures of 873K (in the range 4.0-4.4 MPa m1/2), similar values


52

of elastic modulus (378-358 GPa) and a lower mass decrement rateVn (5.7-7.8 µm/h) in relation to pure alumina. Analysis of results indicates a somewhat higher fracture toughness (by approx.10%) for ceramic composites containing TiC submicro powders than for those containing TiC nano powders (<130 µm). The microstructure of the titanium carbide-reinforced ceramic composites with an alumina matrix is presented in Figures 3 and 4. a)

b)

Fig.3. SEM micrographs of a surface of the ceramic composites: a) Al2O3-ZrO2-TiCsubmicro (B), b) Al2O3-ZrO2-TiCnano (B1). a)

b)

Fig.4. SEM micrographs of a fracture surface of the tested composites: a) Al2O3-ZrO2-TiCsubmicro (B), b) Al2O3-ZrO2-TiCnano (B1). The values of KIC(ET) at 873K are somewhat lower in comparison to those of KIC at room temperature. A small reduction in fracture toughness is observed at elevated temperature (873K) in comparison to that at room temperature in the range of 2-6% for both the Al2O3-ZrO2-TiCnano and the Al2O3-ZrO2-TiCsubmicro composites. Fine particles of zirconia and of titanium carbides, homogeneously

ZrO2


53

distributed in the alumina matrix, are visible in the micrograph of the Al2O3-ZrO2-TiCnano composites (Fig.4). Results of a cutting test carried out on the titanium carbide-reinforced ceramic composites with alumina matrix reveal a more than two-fold increase in tool life (Tmean=24 min) for the Al2O3-ZrO2-TiCsubmicro composites (B) at the accepted wear criterion VBB = 0.30 mm in comparison to the Al2O3-ZrO2-TiCnano composites B1 and B2 (for which tool life achieves values of 9 min and 11 min respectively). The results of the presented investigations allow rational use of existing ceramic tools.

4. Conclusions

• Titanium carbide-reinforced alumina-zirconia composites exhibit: high Vickers hardness, sufficient fracture toughness (critical stress intensity factor KIC is increased to 4.6 MPa m1/2), a high elastic modulus and a higher wear resistance in comparison to pure alumina.

• Analysis of the results indicates a somewhat higher fracture toughness (by approx. 10%) for ceramic composites cotaining TiC submicro powders than for those containing nano powders (<130 µm).

• A small reduction in fracture toughness is observed at elevated temperature (873K) in comparison to that at room temperature in the range of 2-6% for both Al2O3-ZrO2-TiCnano and Al2O3-ZrO2-TiCsubmicro composites.

• Preliminary industrial tests confirm the high cutting performance of titanium carbide-reinforced alumina-based composite tool ceramics.

• Cutting test results indicate a more than two-fold increase in tool life (Tmean) for Al2O3-ZrO2-TiCsubmicro composites in comparison to Al2O3-ZrO2-TiCnano composites.

Acknowledgement

This work was supported by the 2007-2013 Innovative Economy Programme under the EU National Strategic Reference Framework, priority axis 1, section 1.1.3 No UDA-POIG.01.03.01-12-024/08-00.


54

REFERENCES

[1] Handbook of Ceramic Composites. Edited by Narottam E Bansal, NASA Glenn Research Center, Kluwer Academic Publishers, Boston / Dordrecht / London, 2005.

[2] Fortulan C.A., Souza P.D.F., Mat. Research, 1999,2,3, p.205. [3] Y.G. Liu, Y. Zhou, D.C. Jia, Q.C. Meng, Y.H. Chen. Scripta Materialia,

2002, 27, p. 63. [4] Szutkowska M., Journal of Materials Processing Technology, 2004, 153-

154, p. 868. [5] Kee-Do Woo, Byung-Ryang Kim, Eui-Pyo Kwon, Duck-Soo Kang, In-Jin

Shon, Ceramics International, 2010, 36, p.351. [6] Dobrzański L., Materiały inżynierskie i projektowanie materiałowe. WNT

Warszawa, 2006. [7] Fett T., Munz D., J. Am. Ceram. Soc., Vol. 75, No 4 (1992), p.958. [8] Fett T., Eng. Fract. Mech., Vol. 40, No 3 (1991), p.683.

Nanocomposites in the ZrO2-Al2O3-WC system

55

NANOCOMPOSITES IN THE ZrO 2-Al 2O3-WC SYSTEM

ZBIGNIEW PĘDZICH

1, WOJCIECH MAZIARZ2

Abstract

Particulate composite in the Al2O3-ZrO2-WC carbide system shows promising properties in structural applications especially as wear resistant parts. Despite of mechanical properties of constituent phases, its properties depends on the residual stress state, the microstructure and the interphase grain boundaries state. Paper presents the results of investigation of high-resolution transmission electron microscopy (HREM) on the different type grain boundaries structure and its chemical composition.

1. Introduction

Fabrication of the dense composite materials using typical powder compacting technology and following sintering process meets many barriers connected with each step of this technology. Application of submicro-or nanosized powders increases difficulties occurring during manufacturing. The first task for applied technology is to assure the uniformity of different phase distribution in prepared composite powder. The process must eliminate the possibility of one phase grains aggregation. If it is performed with success, sintering could provide to homogeneous composite body. Usually it is profitable for sintering process to avoid of chemical reactions, nevertheless sintering must be provided at temperatures, which assure the effective mass transport in sintered phases. 1 AGH – University of Science and Technology, Faculty of Materials Science and Ceramics, Department of Advanced Ceramics, 30 Mickiewicza Av., 30-059 Krakow, Poland, e-mail: [email protected] 2 Institute of Metallurgy and Materials Science of Polish Academy of Sciences, 25 Reymonta Str., 30-058 Krakow, Poland


56

Additionaly, after sintering process, during cooling, in dense multiphase bodies residual stresses could be created due to thermal expansion coefficient mismatch [1, 2]. This work shows the results of detail investigation of the fine-grained microstructure of three-phase composite containing comparable amount of each phase. The special care was put to describe grain boundaries between the grains of different phases (i. e. alumina-carbide or zirconia-carbide).

2. Experimental

Al 2O3-ZrO2-WC composite powder was prepared from commercial powders (α-alumina: TM-DAR, Taimei Chemicals; yttria stabilized zirconia: 3Y-TZ, Tosoh and tungsten carbide: Baildonit) by wet (ethanol) mixing in attritor. The volume proportion of each phase was 1:1:1. Sintering of composite bodies was conducted in the Thermal Technology hot-press device at 1650ºC (with 45 min. soaking time) in argon atmosphere under 25 MPa pressure. Density of sintered bodies was 8.46 g/cm3 (98.5% of theoretical value). The structure investigations were performed in Tecnai G2 FEG microscope dedicated for high resolution (HREM) observations. The thin foils were prepared by cutting thin slices of compacts using diamond sow and followed dimpling and ion milling using Gatan equipments. Finally the samples were covered by carbon thin layer in order to assure their conductivity. The chemical analyses both point and line scans were performed in the HAADF/STEM mode using a energy dispersive X-ray spectrometry (EDS).

3. Results

Fig. 1 presents a set of high angel angular dark field (HAADF) image and corresponding elemental mapping images recorded in the STEM mode on a marked area.


57

1 µµµµm

HAADF

1 µµµµm1 µµµµm

HAADF

AlK ααααAlK αααα

ZrK ααααZrK αααα

WLααααWLαααα

Fig. 1. HAADF image and corresponding elemental mapping images recorded along marked area.

The obtained contrast is due to a different atomic number of elements (Z2-contrast) and it allows to analyze a particulate phases. Taking to the account a distribution of Al, Zr and W, we can calculate the mean grain size of Al2O3, ZrO2, WC phases, respectively. In this purpose the image analyze software Gatan Digital Micrograph™ 3.10.0 was used. The procedure of analyze was as follow: the elemental mapping images of Al, Zr and W were filtered applying ”maximum” filter, then from obtained binary images the histograms of gray level were created and determination the threshold for maximum intensity of particulate phases were performed. Finally, we obtained the images of particulate phases


58

and measured values of their mean grain size and filled area (Fig. 2). The mean grain size of investigated phases were as follow: Al 2O3 – 330 nm, ZrO2 – 280 nm and WC – 340 nm, and relative area friction: 35.8, 30.7, and 33.5%, respectively. The obtained data were in a good agreement with these used as initial conditions for preparing of compacts (size of powders and volume proportion of each phase). Calculated mean grain size values for composite components are comparable with these calculated using much smaller image magnifications presented in [3].

Al2O3 - 35,8% ZrO2 - 30,7% WC - 33,5%

Fig. 2. Results of images analyze showing distribution of particulate phases in the area indicated in Figure 1.

In order to analyze the grain boundaries between particulate phases a line scan analyses were performed for several grains. Fig. 3 presents result of line scan analyse between WC and Al2O3 grains performed along the marked line. One can see that both Al and W concentration is changing very sharp close the boundary, it means that no diffusion of this elements occurs. Concerning concentration changes of O, and C, is difficult to say about their diffusion, due to contamination effect, limit of detection of light elements by EDS method, thickness of sample and finally necessity of covering by the carbon nonconductive ceramic sample. Additionally, there is no information about the thickness of grains, therefore the overlapping of signal from different grains can occurs.


59

Fig. 3. HAADF image of composite microstructure and result of line scan analyze across the WC and Al2O3 grains boundary.

The observations of grain boundaries were performed in atomic scale also. Fig. 4 presents the high-resolution transmission electron microscopy (HRTEM) showing grains boundary between WC and Al 2O3.

5 nm

HREMAl 2O3

WC

Fig. 4. HRTEM image and corresponding FFT and IFTT images showing the grains boundary between WC and Al2O3.

One can see the sharp boundary between [001] and [451] oriented grains of WC and Al2O3 phases, respectively. The inverse fast Fourier transform (IFFT) images showed that Al2O3 is almost free from structural defects, whereas plenty of dislocations and stacking faults on 010 planes can be distinguish in WC grain.

5 nm

90°

FFT

IFFT

WC [001]

010 010

d=3.0 Ǻ

d=3.0 Ǻ

5 nm

87°

FFT

IFFT

d-110=4.1 Ǻ

d0-14=2.6 Ǻ

Al2O3 [451]


60

Fig. 5 presents a result of line scan analyse between WC and ZrO2 grains performed along the marked line.

100 nm

HAADF

100 nm100 nm

HAADF

Fig. 5. HAADF image of composite microstructure and result of line scan analyze across the WC and ZrO2 grains boundary.

One can see that signal originating from Zr increases in measurement range form 25 nm up to about 110 nm reaching maximum inside the ZrO2 grain. This behaviour can be explain as both overlapping of the WC and ZrO2 grains or some diffusion of Zr into WC grain. However taking to the account character of signal originating form O (almost the same as in case of Zr) most probably this first one explanation is correct. The above assumption that the grain boundaries in Al2O3-ZrO2-WC composite are free from additional phases was proofed by high resolution observation. Despite of chemical reaction, during high temperature treatment, some chemical elements segregation could occurs. In the WC grain the W/C and W/O ratio was changing continuously. It suggests that during sintering oxygen diffuse into the carbide structure. Paralelly, carbon changes its content in carbide grain what has to change locally WC stoichiometry. It was previously observed in [4].

WC

ZrO2


61

5 nm

5 nm 5 nm

5 nm

110

110

100

WC [001]

ZrO2 [111]

011

101

Al2O3 [451]

104

111hex-Al 2O3t-ZrO2

L12-WC

Fig. 6. HRETM image showing a triple point between three phases, corresponding FFT and IFFT images. Fig. 6 presents HRETM image showing a triple point between three phases. Each type of interphase boundaries is represented Al2O3-ZrO2, Al2O3-WC and ZrO2-WC. The corresponding fast Fourier transforms (FFT) allowed to identify the crystallographic structure of coexistence phases. The tetragonal ZrO2 with [111] orientation, L12 WC with [001] orientation and hexagonal Al2O3 with [451] orientation were identified respectively by measurement of distances and angles on corresponded FFT’s images. The phase boundaries areas did not show additional phases inside.

4. Summary

Investigations showed that the composite microstructure consisted of sub-micrometric grains of each phase randomly and homogeneously distributed. The grains boundaries were distinct and clear. This confirms high purity of used starting powders and a lack


62

of chemical reactions between constituent phases at applied sintering conditions. HREM observations of interphase boundaries and the corresponding fast Fourier transforms (FFT) and inverse fast Fourier transforms (IFFT) allowed to identify the crystallographic structure of coexistence phases in boundaries area. Local changing of the chemical composition of WC phase in contact with zirconia grains was detected. However, its nature should be elaborate in details, distinct changes of O/W and C/W ratios suggest that in the area of WC/zirconia interphase boundary chemical changes occurred. WC in contact with alumina was stable. There was no distinct chemical changes on the WC/alumina interphase boundary.

Acknowledgements

The work was supported by the Polish Ministry of Science and Higher Education under project number 11.11.160.364.

REFERENCES

[1] Z. Pędzich, G. Grabowski, W. Maziarz, Wybrane kompozyty ceramiczne o fazach ciągłych – analiza stanu naprężeń i ich wpływ na właściwości, Materiały Ceramiczne/Ceramic Materials, 60 [4] (2008) 258.

[2] Z.Pędzich, G.Grabowski, Wpływ rodzaju wtrąceń na stan naprężeń, wytrzymałość i niezawodność kompozytów ziarnistych na osnowie tlenku glinu, Kompozyty (Composites), 9 [2] (2009) 149.

[3] Z.Pędzich, Zużycie ścierne materiałów kompozytowych na osnowach tlenków glinu i cyrkonu w różnych środowiskach pracy, Materiały Ceramiczne (Ceramic Materials) 61 [2] (2009) 15.

[4] K. Haberko, Z. Pędzich, J. Piekarczyk, M.M. Bućko, W. Suchanek, Tetragonal Zirconia Polycrystals Under Reducing Conditions, in Third Euro-Ceramics, vol. 1, Processing of Ceramics, P.Duran, J.F.Fernandez (Ed.), Faenza Editrice Iberica S. L., Spain, 1993, 967.

TEM investigation of matrix SAFFILTM fiber interfaces in aluminium alloys based composites

63

TEM INVESTIGATION OF MATRIX - SAFFIL FIBER INTERFACES IN

ALUMINIUM ALLOYS BASED COMPOSITES

JERZY MORGIEL1, JACEK KACZMAR

2, MAŁGORZATA POMORSKA

1, KRZYSZTOF NAPLOCHA2

Abstract

The microstructure of metal matrix composites of AA2024, AA 6061 and EN AC 44300 alloys with 10 vol. % Saffil fibers were investigated using transmission electron microscopy. The fibers in the preform were stabilized using liquid glass binder and fired at 800oC resulting in formation of amorphous silicon oxide joints. The preforms infiltration with AA2024 leads to reaction of liquid metal with SiO2 binder substituting it with fine-crystalline mixture of MgO, Θ and silicon. Similar operation with AA6061 leads to binder substitution with porous amorphous Al2O3 and MgO. Only squeeze casting of EN AC 44200 with Mg<0.6 % left most of the binder and allowed to obtain practically clean Saffil/ matrix interfaces.

1. Introduction

The aluminium alloys found already wide application in automotive industry for production of engine blocks, piston heads and others, but stronger and stiffer materials would be very much welcomed. The experiments with reinforcing of high strength aluminium alloys, like AA6061, 2024, 7075 or AlMgCuAg with up to 20 vol. % of Al 2O3 Saffil fibers showed roughly 20% of UTS increase already at ambient temperatures [1-4]. On the other hand, a composite based on AC 8B casting alloys just reproduced the matrix maximum hardness [2]. Simultaneously, both groups of composites retain their ambient temperature strength by ~100oC higher, than their matrixes. 1 Institute of Metallurgy and Materials Scince, Polish Academy of Scinces, 25 Reymonta St., 30-059 Krakow, 2 Institute of Production Engineering and Automation, Wrocław Uniwersity of Technology, Łukasiewicza 3/5, 50-371 Wrocław


64

The reason of different response of matrixes to the same reinforcing Saffil fibers, as well as further optimization of properties of these composites, could be obtained by finding the detail nature of metal/ceramic interfaces in these materials. The problems with production of such composites were overcome partly by pressure infiltration and partly by wetting enhancement with alloying additions, like magnesium, routinely introduced into commercial aluminium alloys. However, switching from the Al – Al 2O3 system in to AlMg – Al2O3 causes that reactive infiltration takes over. In that case, at the metal/ceramic interface an MgO oxide and at 4<Mg<8 wt. % an even more brittle MgAl2O4 spinel may form [1]. The situation in that area is worsened by abundance of oxygen from SiO2 binder easily dissolved in contact with liquid metal containing magnesium. The observations of Al4Cu1Mg0.5Ag matrix with Saffil fibers bonded with SiO2 showed, that in multi-component alloy the MgAl 2O4 may form along the MgO at very low magnesium additions (~1 wt.%) [3,5]. Similarly, the MgAl2O4 was found on β-Al 2O3 particles strengthening Al1Mg0.6Si0.3Cu (AA6262) alloy [6]. The type of phases formed at the Saffil fiber/matrix interface are important, but their microstructure, as well as presence of voids easily nucleated during such phase transformation might be even more detrimental on l properties of such composites. However, the information on that problem is limited. This project was aimed at investigation of the type and microstructure of phases formed at the Saffil/ metal matrix interface in a series of aluminium alloys used in car industry applications.

2. Experimental procedure

Performs made of Safill fibers (10 vol. %) were stabilized with liquid glass (hydrated SiO2), dried and fired at 800oC. Next, they were squeeze cast with an AA2024, 6061 or EN AC 44200 alloys


65

(Table I). Finally, composite materials were solution heat treated and aged to peak hardness. The microstructure was investigated using Tecnai SuperTWIN FEG 200kV transmission electron microscope with EDAX microanalysis system. Thin foils f were cut using Quanta 3D focused ion beam equipped with Omniprobe lift-out attachment.

Table 1. Additions (wt.%) to alloys used for infiltrating the preforms Alloys Mg Si Cu Mn Fe other

AA2024 1,2-1,8 0,4-0,6 3,8-4,9 0,3-0,9 <0,5 <0,11)

AA6061 0,8-1,2 0,4-0,8 0,1-0,4 <0,15 <0,7 <0,42)

EN AC 44200 0,2-0,6 6,5-7,5 <0,1 <0,3 <0,5 <0,23) 1) Cr <0,1, Ti <0,15, Zn <0,25, 2) Cr 0,04-0,35, Ti <0,15, Zn <0,25, 3) Ni<0,1, Zn<0,1, Pb<0,1

3. Results

Immersing of Saffil fibers into liquid glass covers them with thin layer of binder forming bridges between individual fibers during subsequent forming and firing (Fig.1a). The firing of such preform at high temperature fixes the bridges into permanent form and allows to keep the preform shape during squeeze casting with liquid aluminium alloy. The microstructure observations showed, that rough fiber surface is well covered with the binder, which solidifies in an amorphous form (Fig. 1b).

a)

b)

Fig.1. Scanning (Secondary Electrons) image of preform after firing process (a) and transmission image of SiO2 bridge in-between adjacent Saffil fibers


66

The squeeze casting of such preform with AA2024 alloy caused reaction of liquid metal with the binder substituting it with dense nano-crystalline material (Fig.2a). The local chemical analysis using EDS system indicated, that at areas with bigger amount of binder, i.e. so called bridges the surface of the Saffil fibers covers copper-aluminium crystallites, i.e. probably Θ phase (Fig. 3). The centre of such post-binder area is filled with the mixture of the Θ and silicon crystallites immersed in remnants of SiO2 binder phase, while at their outskirts the MgO oxide forms. However, outside the joints, i.e. where Suffil were covered with thinner layer of binder the latter was totally dissolved and fine crystallites of the MgO precipitated directly on the fibers surface. The infiltration of preform with AA6061 alloy resulted in substitution of SiO2 binder with an amorphous highly porous material (Fig. 4). Aside of porosity, larger voids at the Saffil /metal matrix were also observed. The maps presenting local chemical composition helped to determine, that the amorphous material consist of alumina with traces of silicon oxide (Fig.5). Small amount of SiO2 present inside Al2O3 fiber was introduced into them already at their production stage, i.e. the R.F. grade contain up to 4 at. % Si preventing grain growth. The manganese oxide was located in small amount and only at the border of amorphous areas and matrix. The copper containing Θ type precipitates are usually nucleated on these amorphous areas like the one visible in the centre of HAADF image in Fig.5, or directly on Saffil surface. The same experiments were also performed with EN AC 44200 alloy. In that case, most of the binder bridges withstand contact with liquid metal preserving connections in-between Saffil fibers (Fig. 6). More detailed observations of the binder showed that it contain small petal-type voids, which may indicate either start of dissolution of amorphous SiO2 or just represent the effect of sample preparation with high energetic Ga+ beam (Fig.7). The precipitation of silicon was noted both within matrix and at the fiber surface – some of the within the binder area (Fig.8, 9).


67

Fig.2. Transmission image of products of reaction between SiO2 binder and liquid AA2024 alloy.

Fig.3. Scanning-transmission (HAADF) image and series of maps presenting distribution of O, Al, Si, Cu and Mg obtain from area of reaction between SiO2 binder and liquid AA2024.


68

Fig.4. Transmission image of products of reaction between SiO2 binder and liquid AA6061 alloy.

Fig.5. Scanning-transmission (HAADF) image and series of maps presenting distribution of O, Al, Si and Mg obtain from area of reaction between SiO2 binder and liquid AA6061.


69

Fig.6. Transmission image of SiO2 binder bridges between Saffil in perform force infiltrated with AC 44200 alloy.

Fig.7. Transmission image of Si precipitates at Saffil boundaries within binder in perform force infiltrated with AC 44200 alloy.


70

Fig.8. Transmission image of Si precipitates at Saffil boundaries outside binder in perform force infiltrated with AC 44200 alloy.

Fig.9. Scanning-transmission (HAADF) image and series of maps presenting distribution of O, Al, Si and Mg obtain from Saffil/ AC 44200 matrix interface.

4. Discussion and summary

The squeeze casting of ceramic preforms made of Saffil fibers using liquid AA2024, 6061 or EN AC44200 alloys produced dense


71

metal matrix composites. However, an improvement in mechanical properties of these composites could be gained only after ascertaining a proper connection between these high strength (~2 GPa) nano-crystalline slightly porous (5 – 10 %) of 3 -5 µm diameter Al2O3 Saffil fibers and their matrix. The Vaucher and Beffort [1] indicate, that a proper bonding between Al2O3 fibers and aluminum alloy depends to large extent on amount of magnesium additions deciding on type of oxide phases, like MgO or MgAl2O4 formed as a consequence of the production process. They pointed also, that presence of the spinel forming from 4<Mg< 8 wt. % content is the most detrimental for composite mechanical properties. Even, as this amount is higher than usual magnesium additions in commercial alloys, though presence of SiO2 bonding phase might cause nucleation of MgAl2O4 also in case of such alloys, i.e. like Al1Mg2Cu0.5Ag [3] or Al12SiCuMgNi [5]. Present experiments with AA2024 and AA6061 alloys containing 1.6 and 1 wt. % Mg respectively showed however only presence of MgO precipitates. The EN AC44200 alloy with even lower magnesium was free from any magnesium oxides. The observed differences between literature and experiment are probably connected with varying squeeze casting conditions, like infiltration time, pressure and other. The lack of MgAl2O4 spinel at Saffil fiber/matrix interface, generally considered as a necessary condition for obtaining proper composite strength improvement, is not an only requirement. The microcrystalline mixture of Θ-Al 2Cu, Si and MgO material found in Saffil/AA2024 composite has close similarity to the one with spinel obtained after infiltration Saffil with Al4Cu1Mg0.5Ag [3]. Even, as such a mixed microstructure seems of dubious value for load transfer, though the measurement of the latter indeed showed significant strength increase. However, the porous mixture of amorphous Al2O3 and MgO substituting bonding phase in Saffil/AA6061 composite definitely stand no chance at all at keeping fiber-matrix together. What is more, the voids also present in this area might serve, as good crack initiators. In this situation,


72

the Saffil/AC44200 composite with preserved amorphous SiO2 bonding phase at fiber surface and therefore with clean interfaces should present the highest relative property improvement, i.e. as compared with its matrix. The performed microstructure observation aimed at characterization of Saffil fiber/ aluminum matrix interfaces clearly indicated, that the composite mechanical properties are strongly depended not only on type of phase formed over there, but also on defects like porosity or voids. The decreasing magnesium content in the AA2024, 6061 or EN AC44200 alloys used as matrixes in investigated composites resulted in diminishing or even eliminating formation of MgO at fiber interfaces. However, the above alloys differ also in amount of other alloying additions and consequently in melting temperature, so the magnesium content was just one of parameters influencing above changes at Saffil/matrix interfaces.

Acknowledgements

The work was supported through ”KomCerMet” project (contract no. POIG.01.03.01-14-013/08-00 funded by Polish Ministry of Science and Higher Education) in the framework of the Operational Programme for Innovative Economy 2007-2013.

REFERENCES

[1] Vaucher S. and Beffort O., Bonding and interface formation in metal matrix composites, MMC-Assess Thematic Network, EMPA, TU Wien 1998-2001

[2] Automotive Data Sheet Overview, Honda Motor Company, www.saffil.com/pdfs/automotive/data/mmcprop.pdf

[3] Cayron C, TEM studies of interfacial reactions and precipitation mechanisms in Al2O3 short fiber or high volume fraction SiC particle reinforced Al4Cu1Mg0.5Ag squeeze cast composites, Ph. D thesis, Lausane 2000

[4] Long S., Beffort O., Moret and Thevoz Ph., Processing of Albased MMCs by Indirect Sqeez Infiltration of Ceramic Preforms on a Shot-Control High Pressure Die casting machine, Aluminium 76(2000)82-89


73

[5] Dudek H.J., Wang S., Fibre-matrix interaction in the δ - Al2O3 fibre reinforced aluminium piston alloy, Materials Science and Engineering, A205(1996)180-186

[6] Veeresh Kumar G.B., Rao C.S.P., Selvaraj N., Bhagyashekar M.S., Studies on AA6061-SiC and AA7075-Al2O3 Metal matrix Composites, Journal of Minerals and Materials Characterization and Engineering, 9(2010)43-55

Tribological properties of supersonic sprayed nanostructured WC12Co coatings

74

TRIBOLOGICAL PROPERTIES OF SUPERSONIC SPRAYED

NANOSTRUCTURED WC12CO COATINGS

WOJCIECH ŻÓRAWSKI1

Abstract

WC12Co nanostructured composite was prepared by means of the hypersonic spray process (HVOF). The microstructure and composition of tungsten carbide nanopowder were analyzed by scanning electron microscope (SEM) and transmission electron microscope (TEM). Investigations revealed the size of the original particles to be approximately 50-500 nm. The nanostructured sprayed coating was observed by SEM analysis and investigated by X-ray diffractometer (XRD). A denser coating structure with higher hardness was observed compared to conventional coating with a small amount of W2C, WC1-x, W and some amorphous phase. The coefficient of friction for the HVOLF-sprayed WC-12Co nanostructured coating was four times lower than for the conventionally sprayed WC-12Co coating.

1. Introduction

The study of nanostructured materials has been extended to include coatings processed using the thermal spray technique. The possibility of producing coatings with superior properties (e.g. wear resistance) when compared to the conventional thermal spray coatings currently available opens a wide range of research opportunities for different materials. In thermal spray processes, materials in wire or powder form are fed into a flame, arc or plasma spray gun, where they are fully or partially melted and accelerated in a gas stream toward a substrate to be coated. Nanostructured materials cannot be fed directly into the spray guns. They are deposited in the form of suspensions, or powders whose grains are agglomerated nanocrystals [1-3]. Although the coatings produced by 1 Kielce University of Technology, The Centre for Laser Technologies of Metals, Al. 1000-lecia PP7, 25-314 Kielce, Poland, e-mail: [email protected]


75

thermal spraying possess numerous defects, such as pores and microcracks, the process efficiency is much higher and the costs are considerably lower, compared to those of other processes. It is, therefore, possible to use this technique on an industrial scale. Tungsten carbide is widely used in industry to improve the wear resistance of machine parts. WC-Co thermally sprayed coatings are well established as materials highly resistant to wear in a wide variety of situations [4-6]. Therefore nanostructured WC-Co coatings are of particular scientific interest because of their physical and mechanical properties, which are superior to those of conventional materials. Tungsten carbide coatings were deposited using the atmospheric plasma spray (APS) or D-gun process or the high velocity oxy-fuel (HVOF) process. The high temperature of the APS process caused decarburization of the tungsten monocarbide to W2C and the occurrence of metallic tungsten. These phenomena can be described by the following reactions [7]:

2WC → W2C + C (1) W2C → 2W + C (2)

There was a considerable reduction in the amount of carbon, due to oxidation of the coating in the flame, which can be described by the following reaction:

2C + O2 → 2CO (gas) (3)

All these processes are dependent on the thermal spray parameters, which have a significant influence on the coating’s wear resistance. The D-gun process is a method that yields coatings with excellent properties; it is, however, expensive to operate. The high velocity oxy-fuel (HVOF) process is one of the most popular thermal spray technologies. The HVOF process possesses advantages over APS, such as high particle velocities and a lower temperature of the spray stream, producing a denser coating with reduced formation of detrimental reaction products. Due to its flexibility and superior coating quality, the HVOF process has numerous applications in different branches of industry. Investigations carried out in this paper concern the properties of nanostructured and conventional


76

WC-12Co cermet coatings deposited by the HVOLF method, including microhardness, microstructure, phase composition, coefficient of friction and wear resistance.

2. Experiment

The powder and sprayed coating structure was analyzed using tungsten carbide, WC12Co. The nanostructured tungsten carbide powder (Infralloy S7412) and sprayed coating was denoted as WC12Co-N, whilst the conventional powder (Amperit 519.074) and sprayed structure was denoted as WC12Co-A. In both cases, the grain size ranged from 15 to 45 µm. The grain size and surface morphology of the two powders are shown in Figs. 1 and 2, respectively. These materials are designed to be used in thermal spraying. They are produced by agglomeration and sintering of fine powder grains. They contain a very large number of spherical grains, which results in loose powder grains compared to materials with irregularly shaped grains. a)

b)

Fig. 1. WC12Co-N powder, morphology of; a) grains, b) grain surface


77

a)

b)

Fig. 2. WC12Co-A powder, morphology of; a) grains, b) grain surface.

Figs. 1b and 2b show a distinct difference in grain surface morphology between the two materials. This is a result of the different dimensions of the materials applied to produce the powders. The nanostructured powder grains consist of nanometric and submicron WC crystals in the range of 50-500 nm (Fig. 3) whilst in the production of the conventional WC12Co-A powder, 1 µm grains were applied.

a)

b)

Fig. 3. Nanometer and submicron WC12Co-N powder particles; a) on the powder grain surface, b) in the cross-section of the powder grain

The coatings were deposited using a kerosene-fuelled HVOLF thermal spray gun (TAFA JP-5000). The structure and chemical composition of the powders and the coatings were analyzed using the following scanning microscopes: JSM–5400 with an ISIS 300 Oxford (EDS) microprobe and FEI


78

Nova NanoSEM 200. Their phase composition was studied using a Bruker D8 Advance diffractometer with Co-Kα radiation of wavelength λ = 1.78897 Å. A CSM Instruments ball-on-disc type tribotester was used to determine the coefficient of friction under unlubricated conditions for the sprayed carbide coatings. The ball, made of bearing steel 100Cr6, had a diameter of 6 mm. The testing involved the use of a computer to aid in controlling and registering the action of the friction force in the function of time. The parameters of the tribotester were as follows: radius – 10 mm, load – 1 N, linear velocity v = 0.063 m/s, number of laps – 10,000, total distance - 754 m. A schematic diagram of the operation of the tester is presented in Fig. 4. The coatings of the disc samples were sprayed and subsequently ground and polished for one hour. Roughness of the WC12Co-N coating was Ra = 0.043 µm, and that of the WC12Co-A coating was Ra = 0.015 µm. Microhardness, measured with a Vickers indenter with a load of 5 N, was 1159 HV and 966 HV, respectively, as the average of 10 measurements carried out using a CSM Micro Identation Tester.


The microstructures of the HVOLF-sprayed coatings are shown in Fig. 4.

a)

b)

Fig. 4. Cross-section of hypersonic sprayed coating: a) WC12Co-A, b) WC12Co-N

Metallographic images of the two WC12Co coatings showed that there were some small undeformed tungsten carbide grains


79

embedded in the cobalt matrix. From the EDS microanalysis, it was clear that the coating content was different in each zone. The light-coloured grains in the WC12Co coatings testify to a high amount of tungsten, whilst the dark-coloured matrix is an area with a high content of cobalt and a low content of tungsten. Different sizes of tungsten carbide grains were visible in both coatings. The coating deposited using the nanostructured powder had a finer-grained structure with unmodified nanocrystals. It exhibited lower porosity, despite the fact that the nanostructured powder contained larger grains. The coating produced from the conventional powder, on the other hand, had higher porosity, resulting from the higher porosity of the grains, which can be seen in the metallographic images. Microstructure in bright field (BF) and related electron diffractions from an area with grains of different Co content are shown in Fig. 5. In the area with enhanced Co content, an amorphous structure occurs, and crystallite grains possess a hexagonal WC structure.

Fig. 5. Microstructure of WC12Co-N coating

The microstructure of two WC grains in the WC12Co-N coating and related electron diffraction (BF) are shown in Fig. 6. Grains are surrounded by an amorphous cobalt matrix. The dimensions of WC


80

grains are in the range of 200-500 nm. Electron diffraction (SADP) is in conformity with the hexagonal structure and orientation of WC [102]. On the basis of analysis of the microstructure of WC12Co-N sprayed coating (Fig. 6), it can be concluded that nanocrystal grains, present in the powder, are also present in the sprayed coating. This confirms that the process of hypersonic spraying, with a relatively low spray stream temperature and short dwell time of particles in the

Fig. 6. WC grains in WC12Co-N coating.

stream, does not cause decomposition of WC nanocrystals, despite their very small dimensions. Significant changes in the phase composition of WC, which occur particularly during plasma spraying, substantially reduce the properties of coatings. [8, 9]. The influence of heat on sprayed WC12Co-N powder is well seen in the case of cobalt, in which amorphous structures form around nanocrystal grains of WC. Analysis of the powder diffraction patterns showed that the WC grains in the nanostructured powder were half the size of those in the conventional powder. The diffraction patterns revealed the presence of WC and Co both in the WC-Co powder and in the deposited coating. The new phases that appeared in both coatings, i.e. W2C, WC1-x and W, were attributable to the prior disintegration of WC in the spray stream. The diffraction lines of the phases after spraying were considerably wider. This testified to a significant degree of elastic and plastic deformation, i.e. a high level of energy stored in the form of network defects; this particularly in the case of the Co phase [10]. Figure 7 shows time-dependent changes in the coefficient of friction for both coatings. As can be seen in the two diagrams, the


81

coefficient of friction for the sprayed WC12Co-N stabilized after the first 50 m, with a slight tendency for growth. The changes in the coefficient of friction for the WC12Co-A coating revealed a much higher level of growth and significant oscillations.

Fig. 7. Coefficient of friction for hypersonic sprayed coating: a) WC12Co-N coating, b) WC12Co-A coating.

Conclusions

1. The nanostructured WC-12Co coating possessed nanocrystal grains which were present in the sprayed powder.

2. The influence of heat on sprayed WC12Co-N powder forms amorphous structures of Co which surround nanocrystal grains of WC.

3. The microhardness of the nanostructured WC-12Co coating was higher than that of the conventional WC-12Co coating; average microhardness exceeded 1159 HV – 20% more than that of the conventional coating.

4. The coefficient of friction for the HVOLF-sprayed WC-12Co nanostructured coating was relatively stable and was four times lower than for the conventionally sprayed WC-12Co coating.


82

Acknowledgements

This work was funded by the Polish Ministry of Science and Higher Education Grant No. N503 015 32/2296.

REFERENCES [1] Fauchais, P., Montavon, G., Bertrand, G.: From powders to thermally

sprayed coatings. Journal of Thermal Spray Technology 2010, 19 (1-2), pp. 56-80.

[2] Lima, R.S.; Khor, K.A.; Li, H.; Cheang, P.; Marple, B.R.: HVOF spraying of nanostructured hydroxyapatite for biomedical applications. Materials Science & Engineering A, Volume: 396, Issue: 1-2, April 15, pp. 181-187 (2005).

[3] Bolelli, G.; Cannillo, V.; Gadow, R.; Killinger, A.; Lusvarghi, L.; Rauch, J.; Romagnoli, M.: Effect of the suspension composition on the microstructural properties of high velocity suspension flame sprayed (HVSFS) Al2O3 coatings. Surface & Coatings Technology, Volume: 204, Issue: 8, January 15, 2010, pp. 1163-1179.

[4] Fang, Z. Zak; Wang, Xu; Ryu, Taegong; Hwang, Kyu Sup; Sohn, H.Y.: Synthesis, sintering, and mechanical properties of nanocrystalline cemented tungsten carbide – A review. International Journal of Refractory Metals and Hard Materials, Volume: 27, Issue: 2, March, 2009, pp. 288-299.

[5] Chivavibul, Pornthep; Watanabe, Makoto; Kuroda, Seiji; Shinoda, Centavo: Effects of carbide size and Co content on the microstructure and mechanical properties of HVOF-sprayed WC–Co coatings. Surface & Coatings Technology, Volume: 202, Issue: 3, December 5, 2007, pp. 509-521.

[6] Chen, Hui; Gou, Guoqing; Tu, Mingjing; Liu, Yen: Research on the Friction and Wear Behavior at Elevated Temperature of Plasma-Sprayed Nanostructured WC-Co Coatings. Journal of Materials Engineering and Performance, Volume: 19, Issue: 1, February 2010, pp. 1 – 6.

[7] Liao H., Normand B., Coddet C.: Influence of coating microstructure on the abrasive wear resistance of WC/Co cermet coatings. Surface & Coatings Technology 124 (2000), pp. 235-242.

[8] Bartuli, Cecilia; Valente, Teodoro; Cipri, Fabio; Bemporad, Edoardo; Tului, Mario: Parametric study of an HVOF process for the deposition of nanostructured WC-Co coatings. Journal of Thermal Spray Technology, Volume: 14, Issue: 2, June 2005, pp. 187 – 195.

[9] Hewitt, Stephen A.; Laoui, Tahar; Kibble, Kevin K.: Effect of milling temperature on the synthesis and consolidation of nanocomposite WC–10Co powders. International Journal of Refractory Metals and Hard Materials, Volume: 27, Issue: 1, January, 2009, pp. 66-73.


83

[10] Żórawski W., Burakowski T.: Mikrostruktura nanostrukturalnych powłok węglikowych natryskanych naddźwiękowo. Inżynieria Materiałowa 6/2008, pp.608-610.

Amorphous Ti-Si-C thin film deposited on AISI 316L in low temperature

84

AMORPHOUS Ti-Si-C THIN FILM DEPOSITED ON AISI 316L IN LOW

TEMPERATURE

AGNIESZKA TWARDOWSKA1

Abstract

Dual beam IBAD technique was used for Ti–Si–C films formation on AISI 316L austenitic steel substrates in room temperature, from a single Ti3SiC2 compound target. As evidenced by X-ray diffraction, scanning electron microscopy, and transmission electron microscopy and nanoindentation test, resultant Ti-Si-C films were soft, uniform and dense, amorphous or multiphase, with presence of Ti5Si3 (or Ti5Si3Cx), TiSi and TiSi2 traces in amorphous Ti- and Si – rich matrix.

1. Introduction

To obtain desirable properties of multilayer coatings the selection and sequence of layers is of primary importance, because parameters, structure and function of each one give individual contributions and/ or synergistic effect in total properties of resultant composite coating. Both chemical vapor deposition (CVD) and physical deposition techniques (PVD) are used for multilayer coatings synthesis. CVD techniques are very effective, but usually require high substrate temperature, which is not acceptable for metal or polymer substrates preferred for industrial applications. PVD techniques enable low substrate temperature deposition and in general rely on material transfer from condensed phase evaporant or sputtered target. In ion beam techniques ion bombardment is employed to modify the structure and composition of the substrate subsurface (ion implantation) or to densify and modify the structure and composition of previously deposited films (ion mixing) or to eject atoms from the target surface with subsequent deposition on the substrate surface in film form. The main difference in above 1 Dep. Technology, Pedagogical University, 2 Podchorazych St., 30-084 Krakow, Poland


85

mentioned methods is that the primary interest in implantation and ion mixing is the surface composition, structure and properties modification by atoms remain, while in sputtering- atoms remove from the target surface. Ion beam assisted deposition (IBAD) combines mentioned above processes in one, as shown in Fig.1.

a)

b)

c)

d)

Fig.1. Ion beam techniques in surface engineering: a) ion implantation, b) ion mixing, c) ion sputtering, d) ion beam assisted deposition BAD

Titanium-silicon carbide Ti3SiC2 is the most widely researched representant of ternary, inherently nanolaminated compounds, which stoichiometry is written after Barsoum as Mn+1AX n, where M is early transition metal, A- IIIA or IVA group element, X- carbon or


86

nitrogen, n=1, 2, or 3. For simplification, due to n value, the stoichiometry of MAX phases is often given by numbers alone, and noted as 211 (n=1), 312 (n=2), and 413 (n=3). The crystal structure of MAX phases is hexagonal, related to the interstitial transition metal carbides and nitrides (MX) with insertion of A-element layers (the unit cell of Ti3SiC2 is shown on Figure 2), typical c/a ratio for MAX phases is 5–8 [1]. Ti3SiC2 is a light material, resistant to oxidation and thermal shocks as ceramic, but relatively soft, damage tolerant, machinable, thermal and electrical conductive, as metal [2]. Atypical for ceramic characteristic of this carbide (Tab.1) make it desirable for numerous

practical application: as high temperature constructions, protective coatings, low friction surfaces, petrochemical industry, electrical contacts or optics. Table 1. Selected physical and mechanical properties of Ti3SiC2

Density [g/cm3]

Melting Point [ºC]

Thermal Conductivity

at 20ºC [W•m-1•K-1]

Coefficient of Thermal

Expansion [10-6K-1]

Young Modulus

[GPa]

Vickers Hardness

[GPa]

4,53 >3000 30-45 8-12 300-320 4-6

For titanium-silicon carbide thin films synthesis, chemical vapour deposition methods are mostly used but such process requires temperature of the substrate above 1000ºC, which is too high to use most metal or polymer substrates. To lower Ti3SiC2 synthesis temperature to 300ºC or less, different physical vapor deposition methods (PVD) are applied [3,4], but unsuccessfully so the problem is still unresolved. The aim of this work is to apply IBAD method to form in low temperature Ti-Si-C thin film on AISI 316L substrate, from single Ti3SiC2 target. IBAD was chosen as an effective low

Fig.2. Ti3SiC2 unit cell: a= 3.07 Å, c= 17.67 Å [1]


87

temperature thin film deposition method, its versatility, possibility of process control and excellent film to substrate connection.

2. Experimental Procedure

2.1. Ti-Si-C thin films deposition parameters

The parameters of IBAD process i.e. type of ions and ions initial energy, target composition, etc. was determined by the SRIM 2008 computer code. In result, Ti-Si-C coatings were formed with two Ar+ ion beams of the initial energy ~15 keV. For The angle between “sputtering” ion beam axis and normal to the target surface was fixed at 670. Ti–Si–C thin films were grown on AISI 316L austenitic steel substrates (Fe, 0.02% wt. C, 18% wt. Cr, 10% wt. Ni, 2% wt.Mo). A single Ti3SiC2 compound target was used, produced by the SHS method at the Department of Advanced Ceramics AGH (Krakow, Poland) from stoichiometric mixtures of Ti, Si and C. Steel substrates were prepared by cutting from 2 mm thick plate samples (Goodfellow, UK) using a Struers Minitom cut-off wheel and cBN cutting disc, then mechanically ground and polished using Struers MD system working with diamond grinding and polishing suspensions.

2.2. Films characterization

The microstructure, chemical and phase composition of coated substrates were examined by scanning (SEM, EDS) and transmission (TEM, SAED) electron microscopy methods using a JSM-6460LV with EDS INCAX-act 350 Oxford Instrument and Tecnai G2 F20 (200kV). For TEM analysis, cross-sections of coated substrates were prepared by FIB method. Phase composition of as-deposited samples was analyzed by XRD using Siemens D500 Kristalloflex diffractometer, operated with monochromatic Cu Kα


88

(λ=1,54Å) radiation in grazing incident configuration (α=1˚ and 3˚)

and XRD patterns were collected using a step size of 0,02° over 2θ

angular range of 20° to 125°. The hardness and reduced elastic modulus of coated and uncoated substrates were determined by nanoindentation test. CSM with OM imaging system was used working with a Berkovich-type diamond indenter under load of 10 mN. Five indentations were made in three regions of each sample. Load- displacement curves were recorded and hardness HIT

and reduced elastic modulus EIT were calculated using the Oliver & Pharr method.


3.1. Chemical and phase composition

The chemical and phase characterization of formed coatings was not an easy task due to films thickness and limitation of EDS and XRD methods. EDS analysis accompanied both SEM and TEM observation from the coating surface and cross-section respectively. EDS analysis results taken from the surface of coated substrate at area marked in Fig.4 are presented in Tab.2. Table 2. EDS analysis results taken from marked area (Fig.4) of the surface of Ti-Si-C coating, formed by IBAD method on AISI 316L substrate.

area C [%wt.] Si [%wt.] Ti [%wt.] total

spectrum 1 10.37 21.84 67.79 100

spectrum 2 55.19 19.67 25.14 100

In EDS (EDX) analysis accompanying TEM examination of the coating cross-section Ti to Si concentration ratio was determined as 70 to 30 (in %wt.). Low temperature deposition of Ti–Si–C films (temperature less than 300°C) from compound Ti3SiC2 target effects usually in a nanocomposite or nanocrystalline structure, with the presence of TiCx phase in amorphous matrix, enriched in carbon


89

content [5, 6]. In our films, formed by ion beam assisted deposition, carbon content is lowered. XRD analysis did identify neither TiC nor TiCx. The main crystalline phase was Ti5Si3 which coexisted with TiSi and TiSi2 in amorphous TiSi matrix. The presence of Ti3SiC2 was not evident, due to partial correspondence of registered diffraction patterns with characteristic data for that compound.

3.2. Microstructure

Scanning electron microscopy was used to examine the surface of coated substrates. As shown in Fig.4a, IBAD formed Ti-Si-C coating’ surface was dense, flat and smooth. Some increase in the roughness of the surface was observed with film thickness. TEM observation of formed films were provided in plain view and on coated substrate cross-section prepared by FIB method, cut perpendicular to the coating surface. TEM examination confirmed that coatings were continuous and uniform (Fig.4b). The thickness of the coating was in the range 150 nm to ~1 µm, depending on IBAD process parameters. Film to substrate boundary was continuous and smooth. Selected area electron diffraction pattern, presented in Fig.4c, was taken from the film area and confirmed that examined coating was amorphous.


90

3.3. Mechanical Properties

Vickers hardness and reduced elastic modulus were determined by nanoindentation tests. Berkovich-type diamond, indenter was used working under load of 10mN. Load- displacement curves were collected for coated and uncoated substrates. The Oliver & Pharr method was used for calculation of the hardness HIT and elastic modulus EIT. The hardness and reduced elastic modulus for uncoated substrates were HIT = 4.4 GPa and EIT = 250 GPa, respectively. For coated substrates, calculated hardness was 4 GPa, and calculated reduced elastic modulus EIT was 160 GPa. In comparison to CVD derived Ti3SiC2 Vickers hardness of about 6

a)

b)

Fig.4. Microstructure of Ti-Si-C coating formed by IBAD technique: a) SEM image of coating surface with marked area for EDS analysis, b) TEM image (plain view) with SAED of the coating, c) TEM image of coated AISI 316L substrate cross-section

c)


91

GPa in the 0.98–9.8 N load [7], for polycrystalline bulk Ti3SiC2 with 10%vol. TiCx Vickers hardness was 12.7 GPa, for indentation load of 49 mN and 4 GPa for 9.8 N [8], while calculated Young’s modulus were referred as 283 GPa [8] or even 326 GPa for bulk Ti3SiC2 were referred [9]. Such decrease in the hardness and elastic modulus values was reasonable for amorphous state of formed coatings and was often referred as 50 to 60% of values that respective crystalline state [10].

Summary

Ion beam assisted deposition process parameters allowed to form, in low temperature, continuous, dense and flat Ti-Si-C films on AISI 316L steel substrates, by use of single Ti3SiC2 target. Low temperature ion beam assisted deposition of Ti-Si-C results in full or partial film amorphisation, which is desirable effect for possible film application as electrical contacts [11]. Amorphous Ti-Si-C films formed by IBAD are soft: calculated hardness HIT 2.7 GPa to 5.3 GPa and the average value of calculated reduced elastic modulus EIT 160 GPa.

Acknowledgements

The work was supported by the Polish Ministry of Science and Higher Education under project No. N507 451434.

REFERENCES [1] Jeitschko W., Nowotny H., Monatsh. Chem. 98 (1967) 329. [2] Barsoum M.W., Prog. Solid State Chem. 28, 201 (2000). [3] Palmquist J.P., Jansson U., Seppanen T., Persson P.O.A., Birch J., Hultman

L., Isberg P., Appl. Phys. Lett. 81 (2002) 835. [4] Emmerlich J., Palmquist J.P., Hogberg H., Molina-Aldareguia J.M.,

Czigany Zs., Sasvari Sz., Persson P.O.A., Jansson U., Hultman L., J. Appl. Phys. 96 (2004) 4817.


92

[5] Koutzaki S. H., Krzanowski J. E., Nainaparampril J. J., J. Vac. Sci.Technol., A 19 (2001) 1912.

[6] Zehnder T., Matthey J., Schwaller P., Klein A., Steinmann P.A., Patscheider J., Surface & Coatings Technology 163–164 (2003) 238.

[7] Goto T., Hirai T., Mat. Res. Bull. 22 (1987) 1195. [8] Gao N. F., Miyamoto Y., Zhang D.: Journal of Materials Science 34 (1999)

4385 – 4392. [9] Pampuch R., Lis J ., Piekarczk J ., Stobierski L., Ceram. Inter. 19 (1993)

219. [10] Wachtman J. B., Haber R. A.: Ceramic Films And Coatings, Noyes

Publications Mill Road, Park Ridge, New Jersey, USA, 1993. [11] Eklund P. et al. Thin Solid Films 518 (2010) 1851–187.

Ionic methods of coatings formation for a special applications

93

IONIC METHODS OF COATINGS FORMATION FOR A SPECIAL

APPLICATIONS

BOGUSŁAW RAJCHEL1

Abstract

Carbon based coatings are frequently applied to improve of the physicochemical and mechanical properties of covered tools, parts of machines or artificial medical implants. For all those applications the adhesion of coatings to covered substrate must be especially good. The physicochemical properties of carbon coatings are a function of the sp2/sp3 bonds concentrations. The physicochemical properties and adhesion of carbon coatings are strongly defined by method used to their formation. Especially two ionic methods, IBSD – Ion Beam Sputter Deposition and IBAD – Ion Beam Assisted Deposition, can be applied to formation of carbon coatings with good adhesion and with controlled sp2/sp3 bonds concentrations.

1. Introduction

Beam of ions, can be used for modification of a thin surface layer or for formation of the flux of energetic atoms and ions. For this reasons, are existing two fundamental ionic methods. First is the ion implantation and the second is the IBSD (Ion Beam Sputter Deposition). The first one is used for modification of the surface layer and second one is used for creation of coatings. All another, such as the IBAD – Ion Beam Assisted Deposition, are based on those methods. The main idea of the IBSD and the IBAD techniques is shown on Fig. 1.

1 Institute of Nuclear Physics Polish Academy of Sciences., 152 Radzikowskiego Str, 31-342 Krakow, Poland.


94

IBSD IBAD

Fig. 1. The IBSD (left) and the IBAD (right) techniques for formation of complex coatings.

In the method IBSD, the primary problem is to obtain a flux of atoms and ions, with optimal energetic and angular distributions for forming the coating with expected nanostructure. In this method, a well-defined ion beam is used for bombarding a material with a suitably chosen elemental composition. As a result of binary collisions of ions with nuclei the flux of atoms and ions is formed. This process is called as ion sputtering. The sputtering process was described with details by Eckstein [1].

2. Sputtering

Sputtering is one of the processes occurring during ion beam bombardment of solid surfaces. The sputtering process, taking place during the bombarding at an angle θθθθ by beam of ions with atomic number Z1, mass M1 and energy E0 of an amorphous material composed from atoms with atomic number Z2 and mass M2 is shown in Fig.1. The coating is formed from atoms and ions, sputtered by ions beam into a solid angle dΩΩΩΩ defined by the surface to be covered. Nanostructure and physicochemical properties of coatings, depend on distributions of energies and angles of sputtered products. The thickness of the coating depends on the number of atoms emitted


95

into the solid angle. which is defined by the geometry of this process. This number is determined, by the number of bombarding ions and by the so-called “sputtering coefficient”. This coefficient depends on ions beam parameters and composition of the sputtered material. Sputtering coefficient Y(E0, θ) can be expressed as [1]:

( ) ( ) ( )θθθ fEYEY 0,, 00 == (1)

where

( )f

c

f

−

≈

2cos

0

πθθθ

(2)

The Y(E0, θ = 0) can be calculated as

( ) ( )thEEFEY ,0, 00 ==θ (3)

Where the Eth is a threshold energy for sputtering. The sputtering coefficient can be calculated by formulae [1] – [3], from tables [2] or by computer codes used to simulation of interaction of energetic ions and solid materials. This interaction can be simulated by the Monte Carlo methods or by the molecular dynamics techniques. Frequently, to simulation of interaction of ions and amorphous material, the SRIM computer code [3] is used. The SRIM 2008 code, based on the Monte Carlo mechanism, can be used to calculate of ion ranges and sputtering coefficients. In this paper the SRIM 2008 code was applied to calculate of sputtering coefficients of carbon, titanium and silicon from amorphous Ti3SiC2 bombarded by Ar+ or Xe+ ions at energy of 15 keV. All obtained coefficients are shown in the Table 1.


96

Table 1. Sputtering coefficients [atoms/ions] of carbon, titanium and silicon calculated for amorphous Ti3SiC2 sample bombarded by Ar+ or Xe+ ions at energy of 15 keV.

Ar+ Xe+ Impact angle Carbon Titanium Silicon Carbon Titanium Silicon

00 0.3739 1.08 0.2956 100 0.331 0.9742 0.2999 0.4081 1.21 0.3757 200 0.3737 1.09 0.3397 300 0.4784 1.37 0.4275 400 0.6566 1.88 0.5807 500 0.9509 2.72 0.837 600 1.55 4.34 1.15 700 2.03 5.86 1.80 800 2.50 7.06 2.20

The nanostructure of the formed coating is determined by distributions of energies and angles of atoms after sputtering. Those distributions can be calculated as product of two independent distributions:

( ) ( ) ( )φφ gEgEg ss =, (4)

Distributions of energies and angles can be also calculated in simulation of interaction of ions with materials.

3. Experimental results

The IBSD and the IBAD techniques were used for creation of thin TiXSiCY and carbon coatings. The TiXSiCY were formed on the 316L steel. The carbon coatings were formed on the polyurethane substrates. For formation of the TiXSiCY the beam of Ar+ ions at energy of 15 keV was used to sputtering of the Ti3SiC2 target. The impact angle was 670. For formation of carbon coatings as sputtering target the graphite or aluminium plate were used. All formed coating were investigated by the confocal dispersive Raman microspectroscopy. The Raman spectrum obtained for the


97

TiXSiCY sample is shown on the Fig. 2a. The Raman spectrum obtained for the carbon coating is shown in the Fig. 2b.

TiXSiCY Carbon

Fig. 2. Raman spectra obtained for the TiXSiCY and carbon coatings. For experiment the 532nm laser light was used.

The 3D depth profile of the carbon coating formed on the Al substrate is shown on Fig.3.

Fig. 3. The 2D depth profile of the amorphous carbon (a-C) coatings formed on the Al substrate.

Numerical analysis of the Raman spectrum, collected for the TiXSiCY coating, shown peaks characteristic for the Ti3SiC2 phase. The depth profiles, obtained for the carbon coatings, shown a thin interface sublayer.

substrate

coating

G and D peaks


98

4. Discussion

Sputtering is one of the processes occurring during ion beam bombardment of solid surfaces. The sputtering process, taking place during the bombarding at an angle θθθθ by beam of ions with atomic number Z1, mass M1 and energy E0 of an amorphous material composed from atoms with atomic number Z2 and mass M2 is shown in Fig.1.

5. Summary

The IBSD and the IBAD techniques are suitable for formation of complex, multi-elemental and multi-layer coatings on any type substrate. All coatings formed by those ionic methods have very good adhesion to protected surface.

6. Acknowledgements

The work was supported by the MNiSW project, contract no. Nn507 451434 and by the project SPO WKP 1.4.3, “Nanotechnology for special applications in industry and medicine”.

REFERENCES

[1] Eckstein W., Sputtering yields, Vacuum, 2008, 82, 930-934 [2] Yamamura Y, Tawara H., At. Data Nucl. Data Tables 1996, 62, 149 [3] www.srim.org

nano materials

Documents